Community Variations in lengths of day surgery ... · surgery: implications for the efficiency ofsurgical management MyfanwyMorgan, RogerBeech ... (inguinal hernia repair) and short

J7ournal of Epidemiology and Community Health 1990; 44: 90-105

Variations in lengths of stay and rates of day case

surgery: implications for the efficiency of surgicalmanagement

Myfanwy Morgan, Roger Beech

AbstractVariability in lengths of acute hospital stayand rates ofday case surgery is shown to be acontinuing pattern which occurs bothbetween and within countries. A model ofthe determinants of health service activityis presented and the contribution ofdifferent factors to the observed variationsis assessed. Differences in methods offunding health services are identified as amajor determinant of the between countryvariations, while the within countryvariations largely reflect the influence oflocal differences in facilities and servicesand the organisation of care at a hospitallevel, as well as the independent effects ofdifferences in clinical practice style.The main rationale for advocating a

reduction in length ofstay and increased useof day surgery is to increase efficiency byreducing costs per case while maintainingthe quality of care. These criteria of costs,clinical outcomes and patient acceptabilityare examined in relation to day case surgeryfor an intermediate surgical procedure(inguinal hernia repair) and short staysurgery for cholecystectomy. The precisecost savings are shown to depend on themethods of costing, assumptions made andfacilities employed, while factorsinfluencing the outcomes achieved includethe criteria ofpatient selection, the surgicaltechniques employed, and the adequacy ofpreoperative communication. Barriers tothe more widespread adoption of short stayand day case surgery include practical andorganisational constraints on clinicalpractice at a hospital level, lack ofawareness among clinicians as to how fartheir practices differ from current norms,and clinical barriers raised by surgeonswho do not see short stay policies asadvantageous. Mechanisms to promotechanges in clinical practice styles includeindependent professional audit, peerreview, and involvement of clinicians inbudgeting and resource allocation.Assessing quality requires that attention isgiven to patient acceptability andsatisfaction as well as to the monitoring ofclinical outcomes.

An important feature of health service activity isthe variation which occurs between geographicalareas and between individual clinicians. Thisincludes variations at a primary care level in termsof general practitioner referral rates, prescribing

rates and ordering of diagnostic tests.' 2 In thehospital sector particular attention has been paidto the variations in hospital admission rates,which form a major determinant of health servicecosts. This has identified substantial variations inadmission rates between countries, with rates ofadmission for surgical procedures generally beinghigher in the US than the UK, although there arealso large within country variations.7 This paperfocuses on two other indicators ofhospital activity:the length of acute hospital stay for surgicalprocedures and the use of day case surgery.The existence of variations in health service

activity raises questions as to whether they reflectdifferences in the needs ofthe populations served,or form an indicator of a failure to match needsand resources. This failure may involve aninappropriate under-use of resources among thelow use groups if they fail to receive those servicesdeemed to be appropriate for their health needs.Alternatively, an inappropriate use ofservices mayoccur among the high use groups if they receiveservices deemed to be "unnecessary" in terms oftheir health needs (eg, "unnecessary" surgicalprocedures, or "inappropriately" long lengths ofstay. The matching of needs and resources thushas implications for the quality of health servicesand the outcomes achieved.

In economic terms, the deployment ofresources may be regarded as inefficient to theextent that a given outcome may be achieved atless cost. For example, reducing length of stay orsubstituting day case surgery for inpatientadmission may not affect the outcomes achievedfor individual patients, but nevertheless may leadto considerable savings in terms of the health careresources required in their treatment, and istherefore more efficient. To the extent that suchsavings enable larger numbers of patients to betreated with a given level of resources (hospitalbeds, doctors, nurses, etc), the effectiveness of thehealth care system is increased. Questions of theappropriateness ofresource use are thus central tothe promotion ofgood quality, cost-effective care.

In this paper we first review the trends in lengthof stay and rates of day surgery, together with theprevailing between and within country variations.We then examine the causes of these variationsand consider the implications of more innovativepatterns of care. Finally we consider the ways inwhich changes in clinical practice style may befacilitated to promote greater efficiency.

Trends in length of stay and day casesurgeryLENGTH OF STAYThe length of inpatient stay has shown a steady

Department of PublicHealth Medicine,United Medical andDental Schools,St Thomas's Hospital,London SE1 7EHM MorganR Beech

Correspondence to:Ms Morgan

Accepted for publicationJanuary 1990

on May 7, 2020 by guest. P

rotected by copyright.http://jech.bm

j.com/

J Epidem

iol Com

munity H

ealth: first published as 10.1136/jech.44.2.90 on 1 June 1990. Dow

nloaded from

http://jech.bmj.com/

Day case surgery rates and lengths of stay

decline in most countries since the 1940s. This isillustrated by data for the Oxford region; themedian length of stay in NHS hospitals in 1948-1957 was 9 9 days for appendicectomy and 15-0days for cholecystectomy, but had declined to 7 2days and 10-0 days respectively by 1978-83.8 Datapublished in the Hospital In-Patient Enquiry(HIPE) since 1974 similarly show a continuingdecline in mean length of stay for all specialities

Operative procedure 1975a 1980b 1985c

Tonsils and adenoids 4-0 3-5 3-1Inguinal hemia 7-3 5-7 4-9Appendix 7 9 6-2 5-4Gall bladder 16-3 13-6 12-2Prostatectomy 19-6 14-0 11-2Hysterectomy 14-4 11-8 10-2a DHSS/OPCS Hospital In-patient Enquiry: main tables,1975 (HMSO, 1978)DHSS/OPCS Hospital In-patient Enquiry: summary tables,

1980 (HMSO, 1983)c DHSS/OPCS Hospital In-patient Enquiry: summary tables,1985 (HMSO, 1987).

and age groups in NHS hospitals in England andWales. This decline is illustrated in table I for sixoperative procedures.The resource implications of a reduction in

length of stay can be illustrated by comparing thechanges in the numbers of patients treated andbed-days occupied. In England, 1985, there were206 general surgery admissions per 100 000population compared with only 191 admissions in1975. However the reduction in length of staymeant that on average only 18 504 beds were useddaily for general surgical patients in 1985,compared with 24 348 in 1975.9 10A similar trend of a declining length of stay has

also characterised hospital use in the USA,"although overall length of stay continues to belower in the USA, so the relative gap between thetwo countries has been maintained. For example,in 1985, the average length of stay in the USA for'inguinal hernia repair and appendicitis was 3-0days and 4 9 days respectively'2 compared with4-6 and 5-4 days respectively in England andWales. 10

DAY CASE SURGERYA surgical day case is a patient who is admitted forinvestigation or operation on a planned non-resident basis, and occupies for a period a bed in award or unit set aside for this purpose. The idea ofhospital day case surgery is not new, although itsmodern development dates from the 1950s.Farquharson was an early exponent in Britain. In1955 he reported on a series of 458herniorrhaphies carried out on outpatients underlocal anaesthesia, rather than the thenconventional 10-14 day inpatient stay. Amongthese, only 11 developed complications requiringadmission. He thus advocated the widespread useof outpatient surgery, emphasising thetherapeutic advantages of early mobilisation andits effect in reducing waiting lists.'3 The use ofday surgery expanded slowly in the UK duringthe 1960s, with several surgeons reporting theirresults. 14 15A circular promoting day case surgerywas issued by the Ministry of Health in 1967 andhas since been advocated by a number of other

government reports.16 In 1985 day case surgerywas formally endorsed by the Royal College ofSurgeons of England as an important element insurgical care and a booklet setting out the views ofthe Royal College on the scope ofday case surgeryand the facilities required was issued."7

Procedures recommended as suitable for daysurgery are planned, clean surgical procedureswhich require a total operating time not exceeding30 minutes. This includes a large number ofminor procedures as well as a range ofintermediate procedures, such as herniorrhaphy(simple inguinal, femoral, or epigastric hernia)and varicose vein ligation. Examples ofprocedures suitable for day case surgery in thespecialities of general surgery, orthopaedic andhand surgery, urology, paediatric surgery,ophthalmic surgery, otolaryngology, plasticsurgery, gynaecological surgery and oral surgery,are given in the Royal College of Surgeonsguidelines. 17 As the report notes, the excision of avariety of minor lumps which require only localanaesthetic can equally well be undertaken in an

outpatient department, although it may beadministratively easier to classify them as daycases. At the other extreme are intermediatesurgical procedures which may sometimes betreated as day cases but often require inpatientadmission.The limiting factors in the selection of

procedures as suitable for day case surgery, apartfrom the length ofoperating time, are whether thepostoperative pain and discomfort can be relievedwithout recourse to injected analgesics afterdischarge from hospital, whether any skillednursing care or management of dressings isrequired, and whether the procedure is likely toimpair the patients' independence in respect oftoilet functions. In terms of patient selection, theRoyal College of Surgeons lists a number ofcriteria, including the age and fitness of thepatient, housing amenities, the availability ofanother adult to provide care during the night aswell as the day, availability of a telephone, andliving within one hour's journey of the hospital.The Royal College of Surgeons estimates, forexample, that in practice about a third of adultmales between 18 and 65 with groin hernias can besafely treated on a day case basis; two thirds willbe unsuitable for medical reasons, eg, obesity,diabetes, cardiovascular or respiratory disease, orfor social and geographical reasons.Although day case surgery has been practised

for nearly 40 years and is advocated bygovernment reports and by the Royal College ofSurgeons, its adoption in the UK has been fairlyslow. For example, it is estimated that theproportion of patients admitted as day casesincreased from nine per cent of the acute caseloadin 1974 to only 17"' in 1986,18 whereas the RoyalCollege of Surgeons17 suggested that possibly onethird of general surgical admissions in districtgeneral hospitals might be suitable for day casesurgery. The Operational Research Unit of theNorth East Thames Regional Health Authority,using the Royal College of Surgeons guidelines,also estimated that day case surgery in the regioncould be increased from 17"% to 38%, of all generalsurgery.'9 Estimates by a panel of anaesthetistsand general surgeons of the proportion of 83

Table I Mean durationsof stay (days) for sixprocedures, England1975-85

91 on M

ay 7, 2020 by guest. Protected by copyright.

http://jech.bmj.com

/J E

pidemiol C

omm

unity Health: first published as 10.1136/jech.44.2.90 on 1 June 1990. D

ownloaded from



procedures which could be performed on a daycase basis similarly points to a considerableunderuse of day case surgery in the UK.20Day case surgery is more frequently performed

in the USA than in the UK, although precisecomparisons are difficult. This is partly due toproblems in distinguishing between day casesurgery and other ambulatory surgery. However,it is estimated that between 1979 and 1983utilisation of hospital based ambulatoryprogrammes in the USA increased by 86 5%.2iBy 1985, 81% of hospitals provided outpatientsurgery, with outpatient surgical proceduresrepresenting 28% of total surgical procedures inAmerican hospitals.22An important development in the USA has

been the growth of free standing surgical facilities(ie, outside the environs of a fully equippedhospital). One ofthe earliest free standing surgerycentres was opened in Phoenix, Arizona, in 1970,with the name "Surgicenter".23 By 1985 therewere 459 free standing surgical facilities in theUSA undertaking nearly 800 000 procedures perannum, with the projected figure for 1988 being681 facilities undertaking 1-4 millionprocedures.22 These facilities may beindependently owned, part of a corporate chain,or occasionally associated with a hospital, andperform a limited range of minor surgicalprocedures. An analysis of the case load of thePhoenix Surgicenter showed the most commonprocedure undertaken was dilatation andcurettage, followed by laparoscopy andmyringotomy, adenoidectomy, cystoscopy andeye muscle operations.24 Day case surgery in theUK has been entirely provided within theenvirons of a hospital. Hospital based day surgerymay take three main forms: (1) a selfcontained daysurgery unit, (2) a day case ward, though patientsgo from it to the main operating theatre, and (3)day case beds in standard surgical wards.

Explaining between country variationsAn important factor encouraging the generaldecline in lengths of stay and increase in daysurgery has been the far reaching changes inclinical practice. These consist of changes insurgical techniques, as well as in suture material,anaesthesia, methods of pain control, theavailability of new antimicrobial drugs and typesof implant, etc. Major changes have also occurredin beliefs and practices regarding postoperativemanagement, especially regarding the value ofearly ambulation. Changing views regardingpostoperative care can be illustrated in relation toinguinal hernia repair, for which bed rest in aplaster frame for 6 weeks was recommended byBassini in 1890.25 During World War II theaverage time between a hemia repair and return todepot in the British army was 12 weeks and stillincluded 6-7 weeks in bed.26 Since then earlierambulation has increasingly been followed, withmean length of stay decling to 7 3 days in Englandby 1975 and to 4-9 days by 1985 (table I).Although new technological developments and

changes in professional views as to "good" clinicalpractices have encouraged the decline in length ofstay and the greater use of day case surgery, therate of adoption of these practices has varied

between countries. An important factorcontributing to these between country variationsis the differences in the level of health serviceexpenditure and methods of financing thedelivery of health care. Hospitals in the NHSwork on a fixed budget, which is largelydetermined by the size and age/sex structure oftheir catchment populations. A major concern isthus to keep spending within their total budget.There is little financial incentive to adopt shortstay and day case surgery, because although thesepractices reduce costs per case, they are likely tolead to greater total costs and a possible budgetoverspend ifthey result in a greater throughput ofpatients. These budgetary consequences ofincreasing efficiency are referred to as the"efficiency trap". Another feature of the NHS isthat hospital medical staff and managers are paidas salaried employees, although an element ofperformance related pay for hospital managerswas introduced in 1984 as a means of increasingefficiency. There have thus been relatively fewfinancial pressures or incentives on managers andclinicians in the NHS to achieve importantbudgetary constraints on activity throughreduction in lengths ofstay and higher rates ofdaysurgery.

Pressures to increase throughput in the NHShave arisen from the increasing demands forhospital services and the waiting lists for electivesurgery. This has recently led to proposals for thereform of the NHS, outlined in the White Paper"Working for patients".27 A central feature ofthese reforms is the encouragement ofcompetition as a means of increasing efficiency.Thus while retaining a tax funded health servicewith budgets allocated to district healthauthorities, delivery of health care would bethrough internal markets based on competitionbetween providers. Important aspects of this arethat district health authorities would placecontracts with hospitals for surgical services andwould therefore compete on quality and price. Toimprove their chances of success in the marketplace, hospitals will therefore need to achievegreater efficiency with a view to reducing costs percase and hence increasing their pricecompetitiveness. In addition, since revenue isrelated to contracts, increasing throughput wouldresult in greater revenue, thus removing theefficiency trap. Other proposals for increasingcompetition are the provision for some hospitalsto become self governing, as well as theencouragement ofcompetition betweenNHS andprivate hospitals. However, as Robinson28 notes,"the White Paper provides a blueprint rather thana working model and leaves many questionsunanswered, including how far competition willbe allowed to develop and the form contracts willtake." One issue is whether the proposed changesin financing although reducing costs per case willlead to a rapid escalation of total health care costs,as a result of an increased number of patientsbeing treated. (For detailed critiques of proposedreforms see Bevan et al;29 Kings Fund Institute;30Quam;3i Robinson.28)

In the USA, health care is mainly based on feefor service, with coverage being provided by thirdparty payers. Since both physicians and hospitalsare reimbursed at cost, as a fee for each service

92 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



rendered, this has encouraged high rates ofhospital admissions and procedures performed,and led to a rapid escalation of health serviceexpenditure.32 As a result, various costcontainment and regulatory measures have beenintroduced with the primary concem of reducing"unnecessary" hospital admissions andovertreatment.33 However, these measures havealso influenced lengths of stay and rates of daycase surgery. They include requirements forutilisation review of the patient's need foradmission, appropriateness of services receivedand length of stay, and the more favourablecoverage given by insurance carriers forambulatory surgery compared with inpatientadmission.34 While these mechanisms haveformed general features of cost containmentmeasures in American health care, they haveassumed particular importance in relation togovernment funded programmes. Beginning in1983 reimbursement of hospital charges forMedicare patients (over 65s and handicapped) hasbeen based on a prospective payments system.This sets out a predetermined price for aparticular type of care based on diagnostic relatedgroups, which forms a method of classifying acutepatients according to their diagnostic status intoone of 468 categories. The introduction ofprospective payment based on diagnostic relatedgroups was important in severing the direct linkbetween the provision ofservices and payment forthem, and provided incentives for hospitals to bemore efficient by reducing the lengths of stay andintensity of treatment, since this offered thereward of increased profits rather than reducedrevenue.35. In addition, Medicare encouragedambulatory surgery by providing favourablereimbursement rates for physicians.23 Anotherchange in 1984 was that Professional ReviewOrganisations replaced Professional StandardsReview Organisations to monitor hospital useunder the Medicare prospective payment system.Professional Review Organisations are regardedas agents of both cost containment and qualityassurance and have specific objectives designed toachieve these aims. One of their objectives inreviewing admissions is to identify those who cansafely and effectively be treated on an outpatientbasis and thus shift care from inpatient to

36outpatient settings.These regulatory systems and methods of

payment introduced from the early 1980s appearto have been effective in reducing admission ratesand lengths of stay in the USA. For example, ithas been estimated that for the same case mix,lengths of stay for patients covered by Medicarewere 25% lower in 1985 compared with 1980,while hospital admissions of Medicare patientshave fallen since 1983, although there has beenlittle research examining the health outcomes ofsuch changes.37Another development in the USA which has

been important in encouraging new patterns ofpractice has been the expansion of prepaid healthcare, in which a group of physicians providescomprehensive health services for a fixedprospective per capita payment to a definedpopulation of members. There are various typesof managed care, the most common being someform of Health Maintenance Organisation. By

1987 there were over 600 Health MaintenanceOrganisations with 27-7 million subscribers.38Health Maintenance Organisations, unlike the feefor service system, integrate the financing anddelivery of health care. They encourage thecontrol of costs through operating within a fixedbudget, while physicians often benefit from anysurplus achieved. There are thus incentives todeliver cost effective care, with a need to workwithin a global budget and to offer a service whichis attractive in terms of cost and quality so as tomaintain a competitive position, although theelement of competition between a HealthMaintenance Organisation and other providersvaries between areas. The precise impact ofHealth Maintenance Organisations is difficult toassess. However, it appears that they haveachieved some cost savings, due mainly to lowerrates of inpatient admission and a greater use ofoutpatient and ambulatory care. HealthMaintenance Organisations may also have had awider effect in forcing conventional insurers andproviders to become more efficient by adoptingthese patterns ofpractice to maintain their marketshare.39 4This brief outline identifies at a systems level

some of the ways in which different methods offimancing health care may contribute to thebetween country variations in rates of hospitaluse. While not intended as a comprehensiveanalysis, it suggests that whereas a fee for servicesystem encourages high rates of hospitaladmission and procedures performed (includingnew high cost technologies), the introduction offinancial and regulatory measures and pricecompetitiveness can encourage a reduction oflength of stay and a substitution of day casesurgery for inpatient admission. However,whatever the system ofhealth care, general trendshave been the increase in hospital admission ratesand the steady decline in lengths of stay. Thesetrends have been encouraged by new technicaldevelopments and clinical practices, which haveboth increased the range of treatments availableand enabled favourable outcomes to be achievedthrough a more cost effective use of resources,including a shorter length of stay and greater useof day case surgery.

Within country variationsSo far the pattern identified is of a general trendcharacterised by a decline in length of stay andincreased use of day case surgery, but ofconsiderable variations between countries.Another important feature of hospital usage is thegeographical variations which occur withincountries.The within country variations in length of stay

have been noted in the UK over many years andwere documented by Heasman41 and Logan etal.42 More recently, an analysis of length of stay forappendicectomy, inguinal hernia repair, andcholecystectomy within regions of the UK showedthere to be significant differences between districtsafter allowing for differences in the age distributionof patient populations, and excluding cases forwhich a second procedure was recorded as a proxymeasure of severity.43 For example, 68% ofadmissions aged 45-64 years treated for inguinal

93 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



hernia repair in the three shortest stay districts inSouth East Thames Regional Health Authorityspent four days or less in hospital, compared withonly 26% in the three longest stay districts.Similarly, 55% of cholecystectomy admissionsaged 45-64 years spent eight days or less inhospital in the three shortest stay districts in thesame region, compared with only 13% in the threelongest stay districts. Comparisons of theobserved length of stay with the regional averageshowed that for both procedures it was at least13% above the regional average in the threelongest stay districts, resulting in the occupancyof 2000 more bed days, and at least 13% below inthe three shortest stay districts. These differencesarose from differences in the length of both thepreoperative and postoperative period. However,it is interesting to note that stays of29 days or overaccounted for only 3 5% of cholecystectomyadmissions in the three longest stay district healthauthorities and 1 4% in the three shortest staydistricts. This corresponds with the findings ofother studies that the proportion of patients"blocking" beds in surgical wards is generallyquite low, while the majority of longer staysurgical patients do not have stays exceeding onestandard deviation of the expected stay.445Within country variations in hospital use are

also associated with different forms of healthservice funding. A recent trend in the UK hasbeen the growth in elective surgery undertaken inthe private sector, which now caters for about16% of all elective surgery and over 30% in theThames Regional Health Authorities.46 Williamset al,47 in an analysis of the median length of stayfor eight "marker" operations, showed that lengthof stay was consistently shorter for patients in paybeds in NHS hospitals than for patients in eitherindependent acute hospitals or in NHS publicsector beds.For example, the median duration of stay for

inguinal hernia repair (age group 45-64) in each ofthese facilities was 4-4 days, 6-0 days and 5-9 daysrespectively. For cholecystectomy (age group45-64) the median durations ofstay were 9-7 days,9 9 days and 10-7 days respectively. The patternwas similar for both preoperative andpostoperative stay and for each age groupexamined. Differential use of other hospitals forpart of the period of care did not explain thisvariation, with the exception of patients havinghip replacement operations treated in NHS paybeds, of whom one in seven were transferred topublic sector care postoperatively.

Rates of day case surgery, although still low inBritain, also vary between geographical areas. Forexample, in 1986, the proportion of generalsurgery conducted on a day case basis rangedbetween districts from 60% to 31% in South EastThames, 90% to 42% in Northern, 10% to 35% inTrent and 16% to 39% in Wessex.48 Similarly,Henderson et a149 found in a comparison of daycase surgery in five districts in Oxford Regionduring 1976-85 that the proportion of all femalesterilisations conducted as day cases ranged from<1% to 35%, while dilatation and curettagevaried from 1% to 43%.The observed variations in rates of hospital use

raise questions of the reliability of the data. Thereis some evidence of an underrecording of

inpatient procedures in the Hospital ActivityAnalysis, which has implications for thecalculation of admission rates, as well as somevariability in the recording and coding of specificitems, especially diagnostic data.50 Problems ofthe quality and comparability of data betweenareas are probably greater for day case surgery, inview of the differences between hospitals in theways in which day surgery is managed. Forexample, some procedures, such as gastroscopyand minor surgical excisions, may be treated as anoutpatient procedure in some districts and a daycase in others, depending on medical preferenceand local facilities, while in some hospitals littledistinction may be made between an outpatientand a day case. As Hill5" notes, part of theapparent increase in rates of day case surgery mayalso be accounted for by a redefinition of casesthat were previously counted as outpatients,without a change in management. This day caserecording "drift" is encouraged by the setting oftargets for day case surgery based on percentagesof total case load. However, while acknowledgingthese problems of hospital statistics, they do notfully account for the observed geographicalvariations in hospital activity in the UK.

Despite the shorter length of stay in the USAthere are still marked geographical variations.This was commented on by Chassin1' in a reportprepared for the Office of TechnologyAssessment in which he states that "thisphenomenon is the most striking of all variationsin health service use that have been observed". Henotes, for example, that in 1979 the average lengthof stay for Medicare patients varied from a high of131 days in one New Jersey ProfessionalStandards Review Organisation area to a low of6-0 days in a Californian equivalent. In addition,over the past 15 years hospitals in the eastern USAhave had lengths ofstay that are about40% higherthan for hospitals in the western part ofthe USA, which cannot be accounted for bydifferences among regions in age, sex or racedistributions.

Geographic variations also exist in ambulatorysurgery. For example, in Syracuse, New York, ametropolitan area with a relatively limited supplyof acute hospital beds (3 0 per 1000 population),37% of all hospital based surgery was performedon an ambulatory basis in 1983, which wasconsiderably higher than elsewhere in the USA.52These geographical variations in lengths of stayand rates of day case surgery may again beinfluenced by differences in definition andrecording. As in the UK, however, they areunlikely to be fully explained in these terms.The general picture which emerges is therefore

of national trends towards a shorter length of stayand increased use of day case surgery. Howeverthere continue to be important variations inpatterns of surgical management within as well asbetween countries. At one end of the continum isthe existence of lengths of stay which exceed agenerally accepted norm and are substantiallyabove the regional or national average. At theother end are patterns of care which involveconsiderably shorter stays than the current normand a greater use of day case surgery. Thesefindings raise questions of their causes andimplications for the efficiency of resource use.

94 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



Explaining within country variationsThe various groups of factors influencing lengthsof stay are shown in the figure. One importantissue is whether the relatively long length of stayin some areas can be accounted for bygeographical variations in patients' needs, andhence can be viewed as appropriate or"acceptable", or whether it reflects aninappropriate or inefficient use of hospitalresources.

PATIENTS' NEEDSThe contribution of geographical variations inmorbidity to the observed variations in healthservice activity has been examined most fully inrelation to the variations in hospital admissionrates.53-55 This suggests that the large variationsin admission rates cannot be fully accounted forby differences in the incidence of morbidityamong population groups, and also reflects theinfluence ofdifferences in health service provisionand clinical practice styles.53 56

Specific criteria for selection for day casesurgery under the Royal College of Surgeonsguidelines include the patients' general fitnessand home circumstances. Similarly, the poorhealth state ofdeprived populations, and their lessfavourable housing conditions, can be viewed asproducing the need for longer lengths of stay.However, there is the question of the extent towhich differences in the medical and socialcharacteristics ofpopulations are translated into agreater use of hospital inpatient care.A number of studies have provided some

evidence of the influence of patients'characteristics on their length of stay. Forexample, a census of patients occupying beds inacute hospitals in one region in England showedsome tendency for length of stay to increase withdeclining socioeconomic groups after controllingfor level of bed provision.57 Similarly, an analysisof length of stay within selected diagnostic relatedgroups at a hospital in Boston, Massachusetts,.showed that lengths of stay were significantlylonger for patients of lower socioeconomic statusafter adjusting for age, sex, severity of illness andthe doctors' specialty.58 More detailed studies ofthe patient characteristics associated withrelatively long length of stay have identified oldage, disability, poverty, and a lack of carers as

important factors. 45 However, the influence ofthe social and medical needs of the population onlength of stay often depends on local variations inhealth service provision, including the availabilityof alternative accommodation and the efficiencyof discharge planning, as well as the extent towhich clinicians take account of patients' socialcircumstances in their discharge decisions(figure).

HEALTH SERVICE PROVISIONA key factor identified as contributing to withincountry variations in length ofstay is the influenceof historical differences in the levels of healthservice provision between areas. Of particularimportance is the supply of acute hospital beds,with a high level of supply being associated with alonger length ofstay.59 60 This is generally viewedas reflecting the effects of the level of pressure onbeds in raising or lowering thresholds fordischarge, as well as influencing the efficiency oforganisation at a hospital level. However, thepossibility ofincreasing efficiency and achieving agreater throughput of patients also depends onlevels offinance, and the supply ofother resourcesnecessary to cater for the more intensive use ofhospital beds, including staffing levels, theavailability of operating theatre sessions, andother facilities.11 62Length of stay in an acute hospital bed is

influenced not only by the provision of services ata hospital level but also by the availability ofalternative facilities, including convalescent andnursing homes, to provide continued nursingcare, and by perceptions of the quality of generalpractitioner care in the area. The substantialgeographical variations in the availability and useof alternative residential facilities is illustrated byan analysis of admissions for appendicectomy,inguinal hernia repair, and cholecystectomy intwo regions in England. Overall less than two percent of patients were transferred to convalescenthospitals and nursing homes on discharge fromthe acute hospital. However, in one district ineach of the regions studied, over 200,, of patientswere transferred to another facility thus reducingtheir length of stay in an acute bed.63 Similarly,the availability of appropriate facilities andstaffing for day case surgery encourages asubstitution of day case for inpatient care.

Determinants of length ofstay and rates of daycare surgery

Characteristic of patients addmitted 1eg. Age

Severity and complicationsf. Lrn^rh;eI;t;e :ne ISoilciumorlties anl

I Social circumstanm

Characteristics of health care system

eg. Supply of hospital beds and staffing levelsSupply of alternative in-patient facilities andcommunity servicesMethods of financing health care(Controls on hospital utilisation)

ces Clinical practice style Organisation o

eg. Pre-operative testing on IP basis eg. Bed allocalSurgical techniques and anaesthesia AvailablityPostoperative management Time taker

\ Length of stay norms

Length of stay(rate of day case surgery)

A hospital care

ition proceduresof theatres

n by tests, investigations

(outcomes-clinical, social, economical)

95 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



ORGANISATION OF CARE

The efficiency of organisation of care at a hospitallevel forms a product of the broadercharacteristics ofthe health care system, as well asof local circumstances and administrativearrangements (figure). Organisational practiceswhich affect length of stay include the day andtiming of admission, Friday and Saturdayadmissions leading to a longer stay, as doadmissions after 3 pm,64 while preadmissiontesting has the potential to reduce preoperativelength of stay.65 The length of postoperative stayis dependent on the organisation of discharge;when discharge decisions are made at theconsultant's ward round, the frequency of wardrounds can influence length of stay,66 while delaysmay also occur through the lack of early dischargeplanning, and through problems of arrangingalternative accommodation or services requiredfollowing discharge.67 68 Other organisationalfactors include the turnaround time forlaboratories and x ray departments 62 69 and thescheduling and management of theatres.70.

Assessments of the total amount of bed daysthat are occupied "unnecessarily" as a result oforganisational inefficiencies have been based onreviews of individual patient's needs for hospitalcare. A recent study in the UK was undertaken byAnderson et al,68 based on the review of a cohortof patients admitted to the John RadcliffeHospital, Oxford, over a three week period. Allmedical, geriatric and surgical admissions wereincluded apart from obstetrics, gynaecology, andadmissions ofchildren under 16. Patients' "need"to be in an acute hospital was assessed regularlyusing a bed study instrument. This identified ninecriteria which justified acute hospital care. On thisbasis, only 38% of total bed days were regarded asbeing occupied by patients who had positivereasons for being in hospital. The figures rangedfrom 52%/ of patients in surgical beds to only 22%/in geriatric beds. A major reason for theoccupancy of a hospital bed in the absence ofpositive reasons, as judged using the bed studyinstrument, was the delay in discharge due to theneed to await the consultant's ward round. Theauthors put forward a series of recommendationsfor reducing such delays. These includedincreasing the frequency of ward rounds at whichdischarge decisions are made, delegatingresponsibility for discharge decisions to otherstaff, and the use of discharge planning anddiagnosis related protocols to plan length of stay.A study adopting rather different criteria and

methods ofassessment was carried out by Beech eta17' based on the retrospective case record reviewof a random sample of medical and surgicaladmissions to a London teaching hospital.Admissions were assigned to five categories.These included the category "alternative" inwhich admission to an acute inpatient bed couldhave been avoided through day case procedures,outpatient investigations, nursing home care or arehabilitation facility if such facilities wereavailable. This "alternative" category accountedfor 7-3% of surgical bed days, the majorityconsisting of cases who, on the basis of the RoyalCollege of Surgeons guidelines,'7 could have beentreated as day cases. The authors also assessedwhether a reduction in length of stay was possible

through the better management of stay coupledwith improved organisation of ancillary services,or through transfer to a non-acute unit, were thisavailable. It was estimated that in total 14 80, ofbed days could have been saved in these ways,with the greatest potential reduction beingthrough the better management of stay.

Studies in the USA have similarly identifiedsome scope for the more efficient management ofpatients, despite the systems of review and othercontrols on hospital use.72 73

CLINICAL PRACTICE STYLE

Clinical practice is influenced by current medicalknowledge, the characteristics of the healthsystem, and the organisation of care and facilitiesavailable at a hospital level. However variationsbetween clinicians in their style of practice alsomake an independent contribution to variations inhealth service activity.Wennberg and Gittelsohn74 cite various types

of evidence of the effects of clinical practice styleon rates of hospital use, including the findingsfrom physician migration studies of the impact ofone or two physicians on utilisation, and theeffects that feedback of information can have inchanging clinicians' practice style. However,some surgical procedures, including hip fractureand inguinal hernia repair, show much lessvariation in admission rates than hysterectomies,tonsillectomies and prostatectomies.7 This hasbeen attributed to differences in professionaluncertainty, or the degree of medical consensusregarding the clarity of diagnoses, efficacy oftreatment, and timing of that treatment orprocedure. Where considerable professionaluncertainty exists, there is greater scope forprofessional judgement and choice, and hence fordifferences in clinical practice style.Once patients are admitted, the clinical

practices of individual clinicians may influencepatients' lengths of stay through, for example,differences in the extent to which preoperativeinvestigations are performed on an inpatientbasis, and differences in the postoperativemanagement of patients, including beliefsregarding early ambulation and length of staynorms. The contribution of differences in clinicalpractice style to variations in length of stay, aswith admission rates, is likely to be least for thoseconditions and practices for which there isgreatest consensus regarding surgicalmanagement and operative techniques, and whichtherefore present less scope for individual clinicaljudgement.

In some cases clinical practice style isconstrained by the availability of facilities andservices and the organisation of care at a hospitallevel. This was demonstrated in a study byGriffith et al,66 based on a review of 1086 patientsadmitted for elective hernia repair to eighthospitals in Wessex, 1970-71. There were shownto be highly significant differences in the meanpostoperative stay between the nine consultants(3 5 to 7 9 days), although the hospital meansvaried even more (3 8 to 9 3 days). The meanpostoperative stays for consultants operating atthe same hospital were also similar in five of thesix non-general practitioner hospitals studied. Asthe authors observe: "The results of the present

96 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



study support the conclusion that theadministrative procedures ofthe hospital imposeda similar stay on patients under the care ofdifferent consultant firms at the hospitals."The scope for variations in clinical style is

greatest in relation to the use of innovativepractices, including short stay and day casesurgery. Evidence of variation between cliniciansin similar institutions in their use of day case

surgery was provided by Roos75 in a study ofoutpatient surgery among patients aged 20 yearsor over in the eight largest hospitals in Manitoba,1983-84. She showed there were large differencesin rates of outpatient surgery between hospitals,which held even after adjustment for patientcharacteristics and differences in case mix.However, the range among surgeons was even

greater. For example, in the hospital with thelowest rate ofoutpatient surgery, 18% ofsurgeonsperformed 95% or more of dilatation andcurettage procedures on an outpatient basis,whereas nine per cent of surgeons performed lessthan 25% on an outpatient basis. The onlydifference between caseloads identified was that a

rather higher proportion of patients treated bysurgeons favouring outpatient treatment were

resident in the region. Roos concludes that the"large differences in rates of outpatient surgery(is) likely to reflect physicians' preferences". Shealso notes that rates ofday case surgery were loweramong surgeons born in Britain compared withtheir North American colleagues. Such findingsthus indicate that the use of day case surgery

cannot be entirely explained in organisationalterms, as reflecting differences in facilities andstaffing (although this may itself be directlyinfluenced by clinicians). Another importantinfluence, particularly in terms of betweenhospital comparisons, is the differences in clinicalstyles and in the readiness of individual surgeonsto adopt new forms of surgical management.

Surgical practices can be viewed as undergoinga gradual process of change. Thus at any one timethere are likely to be groups of surgeons

pioneering innovatory surgical techniques andforms of patient management, who may begradually followed by others until such practiceseventually become the norm. As Stocking6 notes,the rate of adoption of new technologies andclinical procedures, as with other innovations, isprobably characterised by an S shaped diffusioncurve. The earliest adopters can be classified as

"innovators" who act as opinion leaders and are

followed in terms of timing by "early adopters"and the "early majority". The "late majority"continue to be sceptical for a long period followingthe early uptake of the innovation, while the"laggards" continue their traditional practice.However, the rate of adoption, and hence theshape of the diffusion curve, varies for differentinnovations and is influenced by the perceivedvalue of the innovation by the potential adopters(clinicians) and by those responsible at variouslevels for allocating resources and managing theservice. Applying this to day case surgery andshort stay policies, the slow rate of adoption ofthese practices can be attributed to their limitedclinical advantages, with their main value beingthe cost savings achieved and freeing ofresources,while the economic incentives for adopting these

practices depends on the methods of financingand regulating hospital use.The existence of geographical variations in

patients' needs and the changing styles of clinicalpractice means that some within country variationin patterns of surgical management is to beexpected at any one time. However, the existenceof lengths of stay which are considerably abovethe average can be regarded as representing aninefficient use of acute beds. This is frequentlyassociated with various types of organisationaldelays, including delays in effecting discharge,which may be due to problems ofmanagement aswell as shortages of staff and other resources. Atthe other end ofthe continuum are small groups ofsurgeons with innovatory patterns ofmanagement involving short stays and high levelsof day case surgery. The desirability of thesepatterns of surgical management requires anassessment of the clinical and social outcomes aswell as the cost savings achieved.The costs and outcomes of day case surgery are

examined paying particular attention to themanagement of inguinal hernia repair, althoughthe issues and assumptions raised are applicablemore widely. Inguinal hernia repair forms anintermediate general surgical procedure for whicha high proportion of cases are regarded as suitablefor day case surgery by the Royal College ofSurgeons of England.17 It comprises 12% ofelective surgical procedures in the UK and is oneof six elective procedures with the longest waitingtimes for surgery.76 In terms of resources it hasbeen estimated on the basis of hospital costs perday that inguinal hernia repair costs the healthservice £15 million in England and Wales and£40million in the USA at 1986 prices.77

Costing studiesReductions in lengths of stay and the substitutionof day case surgery for inpatient admissionreduces hospital costs per case. However there arequestions of the precise cost savings achieved andthe existence of any "knock on" effects, or coststransferred to other caring bodies, such as districtnurses, general practitioners, and home helps.Finally, there may be social costs (or savings)linked to the time patients and their families areabsent from work.

HOSPITAL COSTSThe costs of day case surgery and inpatient carehave been estimated using hospital billing,average costs and marginal costs. Varyingassumptions have also been made as to the effectsof releasing resources, which may be used eitherto generate savings or to increase overall caseload.

In the USA, where direct patient billing on anitem for service basis is the norm, it is possible tocompare the charges which patients pay fordifferent durations of stay. This approach hasbeen used to compare the costs of day caseinguinal hernia repair with inpatient repair, withsavings of 45-70% of the inpatient costs beingreported for day case repairs.7"0 However, thisdifference in the cost figures partly reflects theeconomic stance of the institution, in terms ofwhether it is profit making or not, as well as theinfluence of regulatory bodies, such as Medicare.

97 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



Another approach is to employ averagecostings. An example is a study by Burn,81 basedon a review of the complete range of surgery inSouthampton General Hospital. He listedprocedures which he considered suitable for daycase surgery and by an analysis of operatingtheatre records estimated that only 40% of thethen current potential for day case surgery wasbeing achieved. To assess the cost implications ofan expansion ofday case care, Burn used NHS perdiem total average costs per inpatient day and perday case. Average costs were £L11259 perinpatient day and £54-74 per day case. Basedupon an average length of stay for minor surgerycases of two days, Burn estimated that day casesurgery provided a cost reduction of75%, giving apotential saving to the health authority of £1C 27million if the full potential for day case surgerywas realised. However his use of overall averagecosts which cover different levels and intensity ofservices does not adequately reflect individualpatient cost profiles. In addition, many of thecosts would remain fixed after an expansion ofdaycare, at least in the short term, as the actual level ofresources released, either to achieve savings or toexpand the overall level of care, will dependprimarily upon actual long term changes instaffing levels. In the short run possibly only hotelcosts and non-pay costs such as drugs may besaved. When assessing the level of resourcesreleased within hospsitals from an expansion ofday care, marginal costs rather than average costsare therefore required.A detailed study involving a marginal costing

approach was conducted by Russell et a182 in 1972and compared inpatient and day case care foringuinal hemia and haemorrhoids. The studypopulation consisted of all patients requiringnon-recurrent inguinal hemia and haemorrhoidrepair over a one year period at North TeesGeneral Hospital, Stockton-on-Tees. Patientswere initially screened to remove any that wereinappropriate for the trial because of their age,home circumstances or medical condition.Patients were then randomly allocated to day casesurgery, or to an inpatient stay of five days forhemia patients and six days for haemorrhoidpatients. The day case and inpatient groups werecompared in terms oftheir clinical outcome, socialeffects and costs. Adopting a marginal costingapproach, they estimated that the reducedrequirement for bed days resulting from anexpansion of day care made it possible to close afive bedded ward, leading to savings of £25 percase. A major benefit of day surgery is that byfreeing inpatient beds it provides the opportunityto increase service levels. In this study, five bedscould have been released. They thereforeestimated what had been saved by avoiding theneed to construct a new five bed ward. This led to

savings of £33 per case on hospital services. Byadopting a marginal approach to costing, thisstudy thus predicted smaller potential savingsfrom day surgery than the previous study byBurn.81Beech and Larkinson83 also assessed the

marginal cost savings for a hospital from theexpansion ofday care and reduced lengths of stay,based on general medicine and general surgeryservices in West Lambeth. They determined the

savings from maintaining service levels in terms ofcaseload and standards of care with 15-9% fewerbeds, which it had been estimated could have beenreleased based on a previous case record review.7'Of these bed days released, 4-4% were from theexpansion of day care, and the remaining 11 5%from policies to reduce lengths of stay by theremoval of organisational delays. Using averagecosts they estimated that potential sav'ings ofk2 0million per annum could be achieved if non-patient-related overheads were included, and£C1 4 million per annum if they were excluded.However, on the basis of a detailed analysis of thecosts and staffing levels that would actually vary ifthe efficiency gains were realised, they predictedannual savings of only £214 670. These smallsavings stem from the assumption that hospitaloverhead costs will remain fixed, and with theexception of nursing costs, the costs associatedwith the treatment of patients will be unchanged.Remaining savings were related to "hotel" costs.The authors recognise that the scale of theirestimated savings will vary according to localcircumstances but they argue that theirconclusion that the majority of costs remain fixedwhen beds are closed but services are maintainedis generalisable. However the desire to reducetotal costs may not be the motivation forintroducing day case care. The expansion of daycase surgery also frees inpatient beds, thusreleasing the opportunity to increase total servicelevels.The revenue consequences of increasing total

service levels depend on the system of financinghealth care. In a system of fixed budgets, as hasexisted under the NHS, increasing throughputdoes not usually generate further revenue, withthe result that an expansion ofservices can only beachieved through efficiency savings. In contrast,in a market situation where revenue is directlyrelated to the numbers of cases treated, increasingthroughput is not constrained in this way, whilereducing length ofstay and increasing day surgerywill make a hospital more price competitive andincrease its ability to attract patients. In addition,a greater caseload will mean that fixed andoverhead costs can be apportioned across a greaternumber of patients and hence prices per case canfall.

Costing studies thus point to considerablesavings at a hospital level through the reduction inlength of stay, and more especially thesubstitution of day case for inpatient surgery. Forexample, costs of day case inguinal hernia repairare estimated to be between 30% and 60% ofthose of inpatient admission. However, theprecise cost savings achieved at a hospital leveldepend on (1) the methods ofderiving cost figuresand items included, (2) the assumptionsconcerning the use of the bed days saved andwhether the short or long run implications areconsidered, and (3) the levels of staffing andorganisation of services. For example, a daytheatre with a day case ward or five day wardattached offers much greater scope for staffingeconomies than when day cases are included in themain theatre list. High levels of theatre utilisationand low rates ofunscheduled or cancelled sessionsalso help to spread overhead costs. Differences inlevels of theatre utilisation are reflected in the

98 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



varying hourly cost estimates for operatingsessions. In a recent study of six hospitals in theUK the hourly cost for day surgical patientsranged from £73-88 to £142-42, with an averageof £l107-95.84 Other influences on the costs ofdaycase surgery include whether local rather thangeneral anaesthetic is employed and hencewhether an anaesthetist is required.

TRANSFERRED COSTSSurgical practices involving a short length of stayand greater use of day surgery, although reducinghospital costs per patient treated, may produce"knock on" effects for community services, suchas district nurses and general practitioners.Empirical evidence of these knock on ortransferred costs is fairly limited. In Burn's8lopinion, day surgery rarely places demands onthese services. Conflicting evidence came fromstudies of intermediate surgical procedures byRussell et a182 and Adler et al.85

Russell et al,82 in their comparison of inpatientand day case care for the treatment of inguinalhernia and haemorrhoid patients, estimated thatday case patients required an average of4 18 morevisits from district nurses and 0 5 more visits fromgeneral practitioners. At 1972 prices theyestimated that these services cost £4 per day case,to be set against the savings at a hospital level ofday case surgery of £24 to £33 per patient.Adler et a185 assessed a policy of early discharge

for patients requiring treatment for inguinalhemia and varicose veins in which patients wererandomly allocated into two lengths ofpostoperative stay: 48 hours and 6-7 days. Shortstay patients had lower inpatient costs but highercosts associated with general practitioner, districtnursing and home help services. The total cost ofthese services for long stay inguinal herniapatients was £4-88 and for short stay £9-31, a costdifference of £4 43. For varicose vein patientsthese costs totalled £5-26 for long stay patientsand £11-92 for short stay, a difference of £6-66per case. These higher costs are quite large whenset against savings per case on hospital servicescalculated by Adler et al of £32-64 for inguinalhernia patients and £29-84 for varicose veinpatients. Adler and his colleagues also found thatshort stay patients had a longer period ofconvalescence. They estimated the productionlosses of this longer time to return to normalduties as being £19-40 per patient higher forhernia cases and £4-33 per patient higher forvaricose vein cases.These studies emphasise the importance of

taking account of any transferred costs whenassessing the savings achieved by short stay or daycase surgery. Requirements for nursing care andmedical services following discharge havehowever decreased since the 1970s when thesestudies were undertaken, as a result of changes inthe surgical techniques and methods of sutureemployed. In addition, there have been importantdevelopments in the preoperative preparation ofday case patients which may increase their abilityto manage after discharge. Nevertheless, the issueof transferred costs is often raised and identifies aneed for studies to assess the transferred costscurrently associated with day case surgery forintermediate procedures. There is also a need for

studies costing different forms of organisation ofday case surgery, in terms of the facilities andstaffing levels employed.

THROUGHPUT AND WAITING LISTSThe extent to which day surgery is forming asubstitute for inpatient care or merely increasingthe total surgical throughput has been examinedbased on the experience of a small number ofcentres including Kingston Hospital in Surrey,which introduced a six bedded day surgical unit in1979.86 Three relatively major procedures-inguinal hernia repair, varicose veins and partialmastectomy-comprised 26%o of all procedurestreated in the unit and about seven percent of allsurgical operations. However, the authors notedthat for all operative procedures with theexception of three (partial mastectomy,cystoscopy and inguinal hernia repair amongpatients aged 15-44) increasing day surgery rateswere superimposed on steady or increasinginpatient rates. However, only for varicose veinswas there an increase in patients from outside thedistrict. Adding day surgery thus increased thetotal throughput of patients over a given period.This contrasts with the views of Lagoe andMilliren52 who suggested that in New York Statethe difference between the highest inpatient rates(67-1 per 1000 population) and the lowest (45 6per 1000 population) might be explained by asignificant substitution of inpatient forambulatory procedures. However, it should benoted that the lowest inpatient admission rate wasmore than twice that for England. Thus onlywhen a population is adequately or evenoverprovided can day surgery become a substitutefor inpatient care. In a situation where there arewaiting lists for surgery an important effect ofintroducing day surgery can be to reduce waitingtimes. This has its greatest effect for patientsrequiring elective procedures, including inguinalhernia repair, varicose veins, hip replacement,cataract and haemorrhoids, who often have thelongest waiting times because others with moreacute or serious conditions are added to theoperating list ahead of them.87 88

Clinical and social outcomesThe minor surgical procedures which form themajority ofsurgical day cases are likely to producefew increased clinical risks compared withinpatient care, so long as the fairly well establishedcriteria of patient selection are followed andappropriate care is given. Day case surgery forsuch procedures also appears to be associated withconsiderable patient satisfaction.89 90 However,there are important questions of the clinical andsocial outcomes of day case surgery for inguinalhernia repair and other intermediate procedures,as well as of substantial reductions in length ofstay. The effects of short stay policies areexamined paying particular attention to themanagement of cholecystectomy. This is one ofthe eight most commonly performed surgicalprocedures and is thus responsible for significantnumbers of bed days. Calculations based on theper diem cost of hospital bed days suggests thatthis procedure cost the health service £23 millionin England and Wales and £72 million in the USA

99 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from


100Myfanwy Morgan, Roger Beech

in 1986, representing 13% and 16% respectivelof the total costs for eight common electivprocedures.77Examples of studies evaluating the clinica

and/or social outcomes of day case and short stasurgery for inguinal hernia and cholecystectom'are summarised in table II. These studies havyinvolved controlled trials, with either a randonallocation of patients to care groups or thidentification of matched pairs (Nos 1-5), as welas uncontrolled trials based often on a review opatients treated by a single surgeon (Nos 6-11).CLINICAL OUTCOMESMortality rates are extremely low for most electivesurgical procedures, whether conducted as aninpatient procedure or day case.91 It is thereforenecessary to focus on other indicators ofmorbidity and quality of care. One of the mainclinical outcomes employed is the experience ofcomplications. The general conclusion drawnfrom the studies summarised in table II is thatshort stay and day case surgery has no significanteffect in increasing risks of complications.However, this assessment is often based on themore major complications. A greater incidence ofminor problems, such as persistent oozing of thewound, infected wounds, chest infection, etc,although sometimes described, are often notregarded as detracting from clinicaloutcomes.85 92 93 The relatively small numbers ofpatients generally studied also mean that unlessdifferences between groups are fairly large theyare unlikely to reach statistical significance.The other main measures of clinical outcomes

employed are rates of recurrence andreadmissions. In terms of recurrence for inguinalhernia repair, a review ofoperations performed ona day case basis at one hospital identified 4-2%recurrences at 34 months, defined as "thepresence of an expansile cough impluse".94 Theynoted that their recurrence rate lies within therange for inpatient repairs but planned to adoptthe Shouldice operation based on the experienceof Glasgow95 and Devlin et al96 with a view tolowering their recurrence rate still further.Readmission rates tend to be fairly low among

patients undergoing elective surgical procedures.However readmission rates form an importantclinical outcome among elderly medical patients,with a substantial proportion of readmissionsbeing classified as "avoidable".97 98The general conclusion drawn by these

evaluative studies is that short stay policies andday case surgery for inguinal hernia repair andother selected surgical procedures have littleimpact on clinical outcomes, although in somecases the presence of higher rates of minor shortterm complications was noted. However, shortstay and day case surgery is generally appliedselectively, with these results thus relating toparticular categories of patients. Other importantfactors influencing the outcomes achieved are thespecific surgical techniques employed and the useof local or general anaesthesia,96 99-100 as well asthe experience and expertise of surgeons inundertaking the procedure.

Continuing developments in surgicaltechniques and expertise may mean that morefavourable outcomes are currently being achieved

Iy than were demonstrated by earlier evaluations.re Rather than continuing to focus on comparisons

of day case and inpatient surgery, or short andx1 longer lengths of stay, it is thus important toy conduct evaluations and more informal reviews toy identify optimal clinical practices in the conducte of these forms of surgical management.ne SOCIAL OUTCOMES11 This has mainly been assessed in terms oftf patient's satisfaction with their care. The majorityof patients express a high level of satisfaction,

although several studies showed that a significantproportion of patients undergoing day case

isurgery for inguinal hernia repair and varicoseveins would have preferred inpatient care (tableII). Similarly, a significant proportion of the

fcholecystectomy patients reviewed by Reder etIal93 would have preferred a longer period ofFhospital care. However, the reasons for thisipreference and the strength of the preference are

rarely described. There is also the question of thetrade offs that patients are willing to make wherethere are waiting lists for surgery. For example,the choice between a shorter length of stay and anearly operation or a longer length of stay and agreater waiting time (or between immediate daycase surgery and waiting for inpatient care), mayresult in a preference for the more immediate careoption. Patients' anxiety and preferences will alsobe influenced by their actual experiences of shortstay and day case surgery. This emphasises theimportance of evaluating patients' satisfactionwith different forms of preoperative preparationand of examining particular causes of anxiety anddissatisfaction, with a view to improving themanagement of these forms of care.

In conclusion, the studies reviewed in table IIsuggest there is some, but not a significant,increase in risks of complications for selectedsurgical procedures undertaken on a day caserather than an inpatient basis, or under a veryshort compared with a longer length of stay.Similarly, they show that although patientsatisfaction is generally high, a small proportionof patients undergoing these innovative patternsof surgical management would have preferred alonger stay. The clinical and social outcomesachieved are however likely to vary over time andbetween hospitals, and they are influenced bythree broad groups of factors. These are: (1)clinical, eg, the specific surgical techniques andanaesthesia employed, the skill of the surgeon,operating time, and postoperative care; (2)organisational, eg, the type of day case facilitiesemployed, the adequacy of communication withpatients, relatives, general practitioners andnurses, and provision for admission if required;and (3) patient selection, eg, severity of patients'clinical condition, general fitness andpsychological state, their housing conditions andavailbility of care at home.

In view of the considerable experience whichhas now been gained in the management of daycase and short stay surgery for intermediateprocedures, it is important to review theexperiences and practices ofinnovative sugeons interms of each of these variables to determine howfar a consensus is emerging as to optimal patternsof management.



j.com/

J Epidem

iol Com

munity H


nloaded from


101Day case surgery rates and lengths of stay

Table II Studies examining clinical and social outcomes associated with reduction in LOS and/or day case surgery for selected surgical procedures

Authors Study design Clinical criteria Social criteria Results

1. Russell et al Random allocation of 132(1977)82 selected inguinal hernia and

63 haemorrhoids patients to 8hours postop LOS, and 5-6days postop LOS

2. Adler et al Random allocation of 105(1978)85 inguinal hemia and 119

varicose veins patients aged18-64 h to 48 postop and6-7 d postop

ComplicationsReadmission rateLength of convalescence

ComplicationsRate of recurrenceLength of convalescence

Preferred LOSTime off work by familymembers

Satisfaction of patient andfamily with LOS

No significant differences incomplications after herniarepair but greatercomplications among shortstay than longer stayhaemorrhoids. Significantproportion of day cases wouldhave liked overnight stay. Noother significant differences insocial measures

No significant differences inmajor complications or inrecurrence rates but longerconvalescence for shorter stay.Significant number of familiesof short stay patients wouldhave preferred a longer staybut no differences for patients

3. Garraway et al(1978)11

Views of patient and carerRandom allocation of 360selected hernia and varicosevein patients to 3 groups:(a) 2 d acute ward, (b) 2 dconvalescent ward, (c) daycase with GP and districtnurse support

10%o of patients and 200ocarers in day case groupexpressed preference forlonger stay

4. Marks et al Retrospective review of(1980)9° hospital admissions and day

case surgery for 10 proceduresin prepaid group practice forperiod 1967-74

5. Reder et al Retrospective view of case(1983)93 notes for cholecystectomy

patients. Patients classified asshort stay (6 d or less) orconventional stay (7-10 d) andmatched for preop LOS;presence of secondarydiagnosis, age and sex. 18pairs identified

6. Pineault et al Patients undergoing tubal(1985)112 ligation, hernia repair and

meniscectomy selected interms of severity of condition,general fitness and age, andthen randomly assigned to daycase or IP case (n = 182)

Provider satisfaction

Complications up to 2 yearsfollowing surgery

Patients view of discomfort24 h after operationReporting of surgicalproblemsComplications recorded incase notes

Patient satisfaction withquality of care

Patient satisfaction with careand attitude to LOS

Patient satisfaction with LOS

Providers satisfied withambulatory care.930o ambulatory patients and880. inpatients satisfied withquality of medical carereceived

Significantly higherproportion of short staypatients thought LOS tooshort.No other significantdifferences between groups

No significant differences inclinical outcome. Over halfday cases thought LOS tooshort and significantproportion would havepreferred IP care

7. Goldbourne andRuckley (1979)92

Review of 1055 patients foringuinal or femoral hernia orvaricose veins selected for daycase surgery. No comparisongroup.

Timing of dischargeComplications assessed atclinic visits or reported bycommunity nurseReadmission rate

Planned discharge delayed for2 50, patients; 1°o patientsreadmitted; 25°o experiencedsome complications, mainlywound problems

8. Cannon et al Review of 104 admissions for(1982)113 inguinal hernia. Identified 54

patients for 48 h postop stayand further 40 for dischargeprior to or on 4th preop day(ie, 94 patients in earlydischarge group)

Timing of dischargeComplications

Patient satisfaction with LOS Planned discharge delayed for4000 of patients, including440° of proposed 48 hourpostop stay. 1000 patientsdissatisfied with time ofdischarge (20o thought theirLOS was too long and 8",, tooshort)

Review of 50 hemia patientsselected for day case surgery.No comparison group

Timing of dischargeComplicationsRecurrence rate

42 of 50 patients retumedhome same day. Twodischarged patients returnedwith complications.No recurrences

10. Moss (1986)114 Consecutive series of 100elective cholecystectomypatients discharged within24 h following operation.No comparison group

19 patients interviewed bynurse

Two patients withcomplications and 3readmitted.Judged successful form of careby surgeon.95t,' of patients satisfied withearly discharge

Consecutive series of 100elective simplecholecystectomies. Dischargeon second postop day acceptedby 77 patients.No comparison group

Timing of discharge andcomplications among 77 shortstay patients

Patient satisfaction with LOS 271',, patients electing forshort stay would havepreferred longer postop stay.660,, of eligible patientsdischarged on second postopday.No readmissions or deathswithin 30 d of operation.

LOS = length of stay; IP = inpatient

9. Rockwell(1982)78

11. Hall (1987)115



j.com/

J Epidem

iol Com

munity H


nloaded from



Changing clinical practice stylesThis review has identified scope for increasing theefficiency of surgical management through agreater use of short stay policies and day casesurgery, as well as the reduction of the relativelylong lengths ofstay in some hospitals which do notappear to be justified by patient needs. This raisesquestions of the barriers to be overcome andrequirements for increasing the efficiency ofsurgical management. One set of barriers consistsofthe constraints on clinical practice that exist at ahospital level and which arise from theorganisation and financing of hospital services aswell as from local conditions, such as the age andsocial circumstances of patients and thedistribution of beds and staffing betweenspecialties. However, there are also questions ofthe factors influencing individual clinical practicestyles and the readiness of clinicians to adoptinnovative forms of surgical management.

Clinicians may often lack awareness as to howtheir practices differ from current norms. Thishas probably been particularly true in the UK,where there has been little feedback to cliniciansof data on which to assess their practice. Howeversmall groups of surgeons have established theirown data base using microcomputers to reviewperformance, with a view to identifying problemsand producing changes in clinical management.In some cases this review has been limited to theactivities of a single firm.101-103 In other casescomparisons have been made acrosshospitals. 104 105

Clinical barriers limiting the adoption of daycase surgery or short stay policies may also arise ifsurgeons do not regard such forms ofmanagement as having any clinical advantages.For example, surgeons frequently question thedesirability of day case surgery for intermediateprocedures, despite the results of evaluativestudies and the endorsement of these practices byprofessional bodies. Reluctance to adopt thesepractices may also reflect the considerableuncertainty experienced as to precisely howinnovative practices should be implementedlocally, as well as a feeling of unease withunfamiliar forms of patient management,including for example the use of local anaesthesiafor inguinal hernia repair. Greer,'06 on the basis ofinterviews with physicians in the United Statesand Britain, suggests that these clinical concernspartly arise from the inadequate way in whichscientific journals deal with clinical needs. Thisarises from a neglect of clinically important data,in terms of information on indicators and clinicaloutcomes, risks and complications, as well asdetails of the specific procedures and facilitieswhich contributed to the outcomes reported.Greer found that of greater importance in thediffusion of new techniques was the informationand confidence gained from contacts with localinnovators and opinion leaders, as well as theopportunity provided by meetings of specialtysocieties to converse with those with "hands on"experience.Other perceived disadvantages which may be

associated by surgeons with day surgery includeits effect on the balance of their workload. In theUK, the adoption ofday case surgery, particularlyfor intermediate procedures, means that surgeons

of consultant grade are often involved inundertaking large numbers of procedures thatwould otherwise be performed by non-consultantstaff on an inpatient basis. As a result, the mostexperienced surgeons are not able to performother more complex or interesting procedures.However, others question this view and argue thatnon-consultant staff can be trained to achieve ahigh level of competence in performing inguinalhernia repair and other intermediate procedureson a day case basis. They also suggest that day casesurgery can be organised so that it does notdisrupt medical student teaching.The measures adopted to achieve changes in

clinical practice styles have varied betweencountries, reflecting the broader characteristics oftheir health care systems. For example, in the USthere has been considerable reliance on regulatoryapproaches and financial penalties for departuresfrom explicit standards. In contrast, in Britain theemphasis has been on encouraging and facilitatingimprovements in clinical management and theachievement of more cost effective care. This hasincluded methods of increasing the involvementof clinicians in resource decisions in the NHSthrough clinical budgeting and resourcemanagement initiatives in which cliniciansparticipate formally in decisions about the use ofresources.'07 Proposals contained in the WhitePaper for influencing clinical managementinclude not only the changes in the financing ofhealth service delivery but also a commitment togiving clinicians a greater role in the managementof hospitals.27 A central role is also assigned tomedical audit in promoting good quality, costeffective care. This involves provision for externalaudit, through the ability of management toinitiate an independent professional audit, if forexample there is cause to question the quality orcost effectiveness of a service, as well as throughthe creation of an independent AuditCommission. However, major emphasis is givento a more informal system of internal audit inwhich clinicians review their own practices. TheRoyal College of Surgeons'08 describes thisreview as a "process by which medical staffcollectively review, evaluate, and improve theirpractice". It is proposed that this form of medicalaudit should be set up in each specialty andinvolve every hospital doctor. Such peer review isregarded as providing a forum for discussion andmay lead to improved education of staff, quality ofcare, outcome of care, and process of care. Whereguidelines have been developed or targets set,these can form important standards by whichperformance can be assessed. However, even inthe absence of explicit standards, medical auditdoes establish a framework for the identificationof variations in process and outcomes, and thediscussion of areas of concern, as well as assistingin identifying ways in which performance can beimproved. The active involvement of clinicians inthe audit process has been shown to have a muchgreater effect in changing physicians' behaviourthan more passive feedback of data.i09Whereas this section has focused on the

achievement of changes in surgeon's practicestyles, such changes also frequently haveimplications for anaesthetists and the compositionand level of workload of nursing staff. It is thus

102 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



important that attention is given to these widerimplications, as well as to the acceptability andsatisfaction with services among patients, whichprovides an important indicator of quality of care.

ConclusionsVariability in length of stay for surgicalprocedures is a continuing pattern that occursboth between and within countries. The betweencountry variation is strongly influenced bydifferences in methods of funding health caredelivery and the incentives or constraints thisprovides for reducing length of stay andsubstituting day case surgery for inpatientadmissions. The within country variations largelyreflect the influence of local circumstances interms of levels of bed provision, the organisationof care at a hospital level and the availability ofcommunity services, as well as the independenteffect of differences in clinical practice styleswhich are a product of the considerable clinicalautonomy that exists. This suggests that so long asclinical decision making affords scope forindividual judgement and choice, some variationin the management of surgical patients will form acontinuing feature of health care.The main rationale for advocating a reduction

in length of stay and increased use of day casesurgery is that this reduces costs per case, withlittle if any reduction in the outcomes achieved,and so enables more patients to be treated with agiven level of resources. However, as this reviewindicates, the precise cost savings achievedthrough a greater use of day case surgery dependon the costing assumptions made and the facilitiesemployed. Only in relation to intermediatesurgical procedures, such as inguinal herniarepair, are there important questions of clinicaldesirability and patient acceptability of day casesurgery, while reductions in lengths of stay onlyraise such issues when lengths of stayconsiderably shorter than the current norms areproposed.

Scientific evaluations suggest that the clinicaloutcomes achieved and levels of patientsatisfaction for short stay and day case surgery donot differ significantly from more traditionalpatterns of patient management. However, thereare questions of the importance attached to anydifferences identified. In addition, the outcomesachieved are likely to vary as a result ofdifferencesin surgical techniques and in the management ofcare, including the criteria ofpatient selection andthe adequacy of the preoperative communicationand preparation of patients in reducing anxiety.Thus whereas previous evaluations haveindicated that such practices do not significantlyreduce the outcomes achieved, there is now a needto compare different methods of organising daycase surgery and to establish optimal patterns ofsurgical management in terms of clinicaltechniques, patient selection and the organisationof care.The relatively slow rate of adoption of day case

surgery in the UK can be attributed to a lack offinancial incentives and regulatory mechanismspromoting this form of patient management. Inaddition, day case surgery carries little intrinsicinterest for the majority of surgeons. This raises

the question of whether a greater specialisationwithin surgery should be encouraged, in view ofthe differing demands involved in undertakingintermediate procedures on a day case basis.Given the choice, many patients in Britain

would probably currently prefer inpatientadmission for intermediate surgical procedures,although choosing day case treatment for minorexcisions. However, evidence from the USAsuggests that the more widespread practice ofinguinal hemia repair as a day case in that countryhas led to its greater acceptance among patients,with many now preferring not to be admitted as aninpatient."10 Thus just as clinical practices changeover time, so also do patients' expectations andpreferences.Mechanisms for encouraging the widespread

adoption of efficient forms of surgicalmanagement will necessarily vary betweencountries, reflecting the differences in theorganisation and funding of their health caresystems. However, it is important to ensure thatfinancial incentives and other mechanismsdesigned to achieve cost savings do not lead to areduction in the standards of care. This requiresthat outcomes are closely monitored. Animportant aspect of this is the review of patientacceptability and satisfaction with care whichmight be monitored on a routine basis as a meansof quality assurance.

We would like to thank our colleagues for helpfulcomments on earlier drafts and especially Tony Swan,Larry Seidl, Gwyn Bevan, Stirling Bryan and BrendanDevlin.This review, including publication costs, was

financed by the Department ofHealth as part ofa largerproject examining variations in surgical practice. Theauthors alone are responsible for the views expressed.

1 Eisenberg JM. Doctors' decisions and the cost ofmedical care.Michigan: Health Administration Press Perspectives,1986.

2 Dowie R. General practitioners and consultants: a study ofoutpatient referrals. London: King's Fund Institute, 1983.

3 Vayda E. Comparison of surgical rates in Canada and inEngland and Wales. N Engl J Med 1973; 289: 1224-9.

4 McPherson K, Strong PM, Epstein A, Jones L. Regionalvariations in the use of common surgical procedures:within and between England and Wales, Canada and theUnited States of America. Soc Sci Med 1981; 15A:273-88.

5 McPherson K, Wennberg JE, Hovind OB, Clifford P.Small-area variations in the use of common surgicalprocedures: an international comparison ofNew England,England and Norway. N Engl J Med 1982; 307: 1310-4.

6 Wennberg JE, Barnes BA, Zubkoff M. Professionaluncertainty and the problem ofsupplier-induced demand.Soc Sci Med 1982; 16: 811-24.

7 Wennberg JE, Freeman JL, Culp WJ. Are hospital servicesrationed in New Haven or over-utilised in Boston? Lancet1987; i: 1185-8.

8 Coulter A, McPherson K. Waiting times and duration ofhospital stay for common surgical operations: trends overtime. Community Med 1987; 9: 247-53.

9 DHSS/OPCS. Hospital inpatient enquiry, 1975 (maintables). London: Her Majesty's Stationery Office, 1978.

10 DHSS/OPCS. Hospital inpatient enquiry, 1985 (summarytables). London: Her Majesty's Stationery Office, 1978.

1 1 Chassin M. Variations in hospital length of stay: theirrelationship to health outcomes. Health Technology CaseStudy 24. Washing DC: Office of TechnologyAssessment, 1983.

12 Department of Health, Education and Welfare. Vital andhealth statistics, series 13. 1985: No. 69. Washington DC,1987.

13 Farquharson EL. Early ambulation with special referenceto hemiorrhaphy as an outpatient procedure. Lancet 1955;ii: 517-9.

14 Stephens FO, Dudley HAF. An outpatient organisation foroutpatient surgery. Lancet 1961; i: 1042.

15 Morris D, Ward A, Hardiside AJ. Early discharge afterhernia repair. Lancet 1968; i: 681-5.

16 Stocking B. Initiative and inertia: case studies in the NHS.London: Nuffield Provincial Hospitals Trust, 1985.

103 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



17 Royal College of Surgeons of England. Commission on theprovision of surgical services:guidelinesfor day case surgery.

London: Royal College of Surgeons, 1985.18 NHS. NHS hospital activity statistics for England 1974-

86. Stat Bull 1987; 5: 87.19 Bates T, Hamm J. Data unto the day patient. Health ServJ

1989; 2 March: 266.20 Gabbay J, Francis L. How much day surgery? Delphic

predictors. Br MedJ 1988; 297: 1249-52.21 Lagoe RJ, Bice SE, Abulencia PB. Ambulatory surgery

utilization by age level. AmjPublic Health 1987: 77: 33-7.22 Riffer J. Can surgicenters stand alone? Hospitals 1986; July

20: 44-8.23 Ermann D, Gabel J. The changing face of American health

care: multihospital systems, emergency centres andsurgery centres. Med Care 1985; 23: 401-20.

24 Lairson D, Baron A, Surint JM, Greenlick MR, Marks SD.Do-not-admit versus inpatient surgery in an HMO:determinants of choice and the implications for medicalcare costs. Health Serv Res 1980; 15: 378-96.

25 Bassini E. Uber die Behandlung des Leitenbruches. ArchKlin Chir 1890; 40:429-39 (Quoted in Morris GE, JarrettPE. J R Coll Surg Lond 1987; 69: 97-9).

26 Edwards H. Inguinal hemia. Br J Surg 1943; 31: 172-85.27 Secretaries of State for Health, Wales, Northern Ireland

and Scotland. Working for patients. (CM555) London:Her Majesty's Stationery Office, 1989.

28 Robinson R. New health care market. Br MedJ 1989; 298:437-9.

29 Bevan G, Holland WW, Mays N. Working for whichpatients at what cost? Lancet 1989; i: 947-9.

30 Kings Fund Institute. Managed competition: a newapproach to health care in Britain. Briefing Paper No. 9.London: King's Fund Institute, 1989.

31 Quam L. Improving clinical effectiveness in the NHS: analternative to the White Paper. Br Med J 1989; 299:448-50.

32 Quam L. Post-war American health care: the many costs ofmarket failure. Oxford Rev Econ Policy 1988; 5: 113-23.

33 Schroeder SA. Strategies for reducing medical costs bychanging physicians behaviour. IntJ Technol Assess 1987;3: 39-50.

34 Feldstein PJ, Wickizer TM, Wheeler JRC. Private costcontainment. NEngl JMed 1988; 318: 1310-4.

35 Bevan G. Reforming UK health care: internal markets oremergent planning. Fiscal Studies 1989; 10: 53-71.

36 Dans PE, Weiner JP, Otter SE. Peer Review Organisations:promises and pitfalls. N Engl J Med 1985; 313: 1131-7.

37 Schramm CJ, Gabel J. Prospective payment: someretrospective observations. N Engl J Med 1988; 318:1681-2.

38 Rayner G. HMO's in the USA and Britain: a new prospectfor health care? Soc Sci Med 1988; 27: 305-20.

39 Luft HS. How do Health Maintenance Organisationsachieve their savings? New EnglJMed 1978; 298: 1336-43.

40 Luft T. Competition and regulation. Med Care 1985; 23:383-400.

41 Heasman MA. How long in hospital? A study in variation induration of stay for two common surgical conditions.Lancet 1964; ii: 539-41.

42 Logan RFL, Ashley JSA, Klein RE, Robson DM.Dynamics of medical care: the Liverpool study into use ofhospital resources. Memoir No. 14. London: LondonSchool of Hygiene and Tropical Medicine, 1972.

43 Morgan M, Paul E, Devlin HB. Length of stay for commonsurgical procedures: variation among districts. Br J Surg1987; 74: 884-9.

44 Seymour DJ, Pringle R. Elderly patients in a generalsurgical unit: do they block beds? Br Med J 1982; 284:1921-3.

45 Coid J, Crome P. Bed blocking in Bromley. Br MedJ 1986;292: 1253-6.

46 Nicholl JP, Beeby NR, Williams BT. Role of the privatesector in elective surgery in England and Wales, 1986. BrMedJ 1989; 298: 243-6.

47 Williams BJ, Nicholl JP, Thomas KJ, Knowleden J.Differences in durations of stay for surgery in the NHSand private sector in England and Wales. BrMedJ3 1985;290: 978-80.

48 Department of Health and Social Security. Performanceindicators for the NHS. London, 1987.

49 Henderson J, Goldacre M, Griffith M, Simmons HM. Daycase surgery: geographical variations, trends andreadmission rates. J Epidemiol Community Health 1989;43: 301-5.

50 Butts MS, Williams DRR. Accuracy of hospital activityanalysis data. Br Med J 1982; 285: 506-7.

51 Hill A. Day cases: what are we counting? Community Med1988; 180-5.

52 Lagoe RJ, Milliren JW. A community based analysis ofambulatory surgery utilisation. Am J Public Health 1986;76: 150-3.

53 Sanders D, Coulter A, McPherson K. Variations in hospitaladmission rates: a review of the literature. London: King'sFund Institute, 1989.

54 Blumberg MS. Inter-area variations in age adjusted healthstatus. Med Care 1987; 25: 340-53.

55 Wennberg JE. Population illness rates do not explainpopulation hospitalisation rates. Med Care 1987; 25:354-9.

56 Morgan M, Maya N, Holland WW. Can hospital use be ameasure of need for health care? J Epidemiol CommunityHealth 1987; 41: 269-74.

57 North East Thames Regional Health Authority. Patientcensus study: social factors analysis. NETRHAManagement Services Report No. 1249, 1983.

58 Epstein AM, Stems RS, Tognetti J, et al. The associationof patients' socio-economic characteristics with the lengthof hospital stay and hospital changes within diagnostic-related groups. N Engl J Med 1988; 318: 1579-85.

59 Feldstein MS. Hospital bed scarcity: an analysis of theeffects of inter-regional differences. Economica 1965; 32:393-409.

60 London Health Planning Consortium. Acute hospitalservices in London: a profile by the London Health PlanningConsortium. London: Her Majesty's Stationery Office,1979.

61 Ro KK. Patient characteristics, hospital characteristics,and hospital use. Med Care 1969; 7: 295-312.

62 Eastaugh SR. Organisational determinants of surgicallengths of stay. Inquiry 1980; 17: 85-96.

63 Morgan M. Variations in length of stay at district level.Chapter 3. In: Ham C, ed. Health care variations: assessingthe evidence. London: King's Fund Institute, 1988.

64 Babaro DM, ShumanLJ, Swinkola RB. An evaluation ofvarious pre-surgical testing procedures. Inquiry 1977; 14:369-83.

65 Cannoodt LJ, Knickman JR. The effects of hospitalcharacteristics and organisational factors on pre- andpost-operative lengths of hospital stay. Health Serv Res1984; 19: 561-85.

66 Griffiths M, Water WE, Acheson ED. Variation in hospitalstay after inguinal herniorrhaphy. Br Med J 1979; i:787-9.

67 Cable EP, Mayers SPJ. Discharge planning effect on lengthof hospital stay. Arch Phys Med Rehabil 1983; 64: 57-60.

68 Anderson P, Meara J, BrodhurstS, AttwoodS, TimbrellM, Gatherer A. Use of hospital beds: a cohort study ofadmissions to a provincial teaching hospital. Br Med J1988; 297: 910-12.

69 Ferrow LC, McColl I, Mackie. Firm, patient and processvariables associated with length of stay. Br MedJ 1978; i:556-9.

70 Robbins JA, DonaldsonLJ. Analysing stages of care inhospital stay for fractured neck of femur. Lancet 1984; ii:1028-9.

71 Beech R. ChallahS, Ingram RH. Impact of cuts in actuebeds on services for patients. Br MedJ 1987; 294: 685-88.

72 Restuccia J. The effect of concurrent feedback in reducinginappropriate hospital utilisation. Med Care 1982; 20:46-62.

73 Sui AL, Sonnenberg FA, Manning WG. Inappropriate useof hospitals in a randomized trial of health insurance plans.NEngl JMed 1986; 315: 1259-66.

74 Wennberg JE, Gittelsohn A. Variations in medical careamong small areas. Sci Am 1982; 246: 100-12.

75 Roos NP. What is the potential for moving adult surgery tothe ambulatory setting? Can Med Assoc J 1988; 138:809-16.

76 Donaldson L, Pugh EJ, Smyth EJ. A census of patientswaiting for treatment or consultation in surgicalspecialities in an English health region. Community Med1989; 11: 13-20.

77 McPherson K. Variations in hospitalisation rates: why andhow to study them. In: Ham C, ed. Health care variations:assessing the evidence. London: Kings Fund Institute,1988.

78 Rockwell E. Outpatient repair of inguinal hernia. Am JSurg 1982; 143: 559-60.

79 Coe RC. Changing methods best way to cut costs. Am MedNews 1981; 24: 5.

80 Flanagan L, Bascom JU. Herniorrhaphies performed onoutpatients' under local anaesthesia. Surg Gynecol Obstet1981; 153: 557-60.

81 Burn JM. Responsible use of resources: day surgery. BrMedJ 1983; 286: 492-93.

82 Russell IT, Devlin HB, Fell M, Glass NJ, Newell DJ.Day-case surgery forhernias and haemorrhoids: a clinical,social and economic evaluation. Lancet 1977; ii: 844-7.

83 Beech R, Larkinson J. Estimating the financial savings frommaintaining the level of acute services with fewer hospitalbeds. London: Department of Community Medicine,UMDS, St Thomas' Campus, 1989.

84 NHS Management Executive VFM Unit. The managementand utilisation of operating departments. London, 1989.

85 Adler MW, Waller JJ, Creese A, Thorne SC. Randomisedcontrolled trial of early discharge for inguinal hernia andvaricose veins. J Epidemiol Community Health 1978; 32:136-42.

86 Haworth E, Balarajan R. Day surgery: does it add to orreplace inpatient surgery? Br Med j 1987; 294: 133-5.

87 Rainey JB, Ruckley CV. Work of a day unit. Br MedJ 1979;ii: 714-7.

88 Lakhani 5, Leach RD, Jarrett PEM. Effect of a surgical dayunit on waiting lists. J R Soc Med 1987; 80: 628-9.

89 Jensen J, Jackson B. Consumers prefer same-day surgery toinpatient care for minor procedures. Modern Healthcare1985; 15: 76-8.

90 Marks 5, Greenlick MR, Hurtado A, Johnson J,Henderson J. Ambulatory surgery in an HMO: a study ofcosts, quality of care and satisfaction. Med Care 1980; 18:127-46.

91 Buck N, Devlin HB, Lunn JN. Confidential enquiry intoperioperative deaths. Association of anaesthetists of GreatBritain and Ireland and Association of Surgeons of GreatBritain and Ireland. London: Nuffield ProvincialHospitals Trust/Kings Fund, 1987.

104 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from



92 Gouldbourne IA, Ruckley CV. Operations for hernia andvaricose veins in a day-bed unit. BrMedJ 1979; ii: 712-14.

93 Reder VA, Fineberg HV, Rosoff C, White LS. Shorterlength of stay for simple cholecystectomy: cost-effectiveness of alternative strategies. Med Care 1983; 21:745-54.

94 Morris GE, Jarrett PE. Recurrence rates following localanaesthetic day case inguinal hernia repair by juniorsurgeons in a district general hospital. Ann Surg Eng 1987;69: 97-9.

95 Glassow F. Short stay surgery (Shouldice technique) forrepair of inguinal hernia. Ann R Coll Surg Lond 1976; 58:133-9.

96 Devlin HB, Russell IT, Muller D, Sahay AK, Tiwari PN.Short stay surgery for inguinal hemia: clinical outcome ofthe Shouldice operation. Lancet 1977; i: 847-9.

97 Graham H, Livesley B. Can readmissions to a geriatricmedical unit be prevented? Lancet 1983; i: 404-6.

98 Williams El, Fitton F. Factors affecting early unplannedreadmission ofelderly patients to hospital. BrMedJ 1988;297: 784-7.

99 Glassow F. Inguinal hernia repair using local anaesthesia.Ann R Coll Surg Lond 1984; 66: 382-7.

100 Devlin HB. Management of abdominal hernia. London:Butterworths, 1988.

101 Dunn DC. Audit of a surgical firm by microcomputer: fiveyears experience. Br Med J 1988; 296: 687-91.

102 Sellu DP, Benjamin IA, Lynn JA. Audit: its effect on theperformance of a surgical unit in a district generalhospital. Hosp Health Serv Rev 1986; 82: 64-9.

103 Irving M, Temple J. Surgical audit: one year's experiencein a teaching hospital. Br Med a 1976; ii: 746-7.

104 Gilmore OJA, Griffiths NJ, Connolly JC, et al. Surgicalaudit: comparisons of the workload and results of twohospitals in the same district. Br MedJ 1980; 281: 1050-2.

105 Gruer R, Gordon DS, Gunn AA, Ruckley CV. Audit ofsurgical audit. Lancet 1986; i: 23-5.

106 Greer AL. The state of the art versus the state of thescience: the diffusion of new medical technologies intopractice. Int J Technol Assess Health Care 1988; 4: 5-26.

107 Wickings I, Coles JM, Flux R, Howard L. Review ofclinical budgeting and costing experiments. Br Med J1983; 286: 575-8.

108 Royal College of Surgeons. Guidelines to clinical audit insurgical practice. London: Royal College of Surgeons,1989.

109 Mitchell MW, Fawkes FGR. Audit reviewed: doesfeedback on performance change clinical behaviour? J RColl Physicians Lond 1985; 19: 251-4.

110 Wantz GE. Ambulatory hernia surgery. BrJ Surg 1989; 76:1228-9.

111 Garraway WM, Cuthbertson C, Fenwick N, Ruckley CV,Prescott RJ. Consumer acceptability of day care afteroperations for hemia and varicose veins. J EpidemiolCommunity Health 1978; 32: 219-21.

112 Pineault R, Constandriopoulos AP, Valois M, Bastion ML,Lance JM. Randomized clinical trial of one-day surgery.Med Care 1985; 23: 171-88.

113 Cannon SR, Ralphs DN, Bolton JP, Wood JJ, Allen A.Early discharge following hernia repair in unselectedpatients. Br J Surg 1982; 69: 112-3.

114 Moss G. Discharge within 24 hours of electivecholecystectomy. Arch Surg 1986; 121: 1159-61.

115 Hall RC. Short surgical stay: two hospital days forcholecystectomy. Am I Surg 1987; 154: 510-5.

105 on M


http://jech.bmj.com

/J E

pidemiol C

omm


ownloaded from


Documents

Community Variations in lengths of day surgery ... · surgery: implications for the efficiency ofsurgical management MyfanwyMorgan, RogerBeech ... (inguinal hernia repair) and short