Costing in the NHS

7/31/2019 Costing in the NHS

1/12

Costing in the National Health Service:from reporting to managing

Christopher S. Chapman and Anja KernImperial College London


2/12

1 | Costing in the National Health Service: rom reporting to managing

Key fndings

Set out more clearly the underlying philosophy o cost analysis and management inocial DH documents.

Develop training or NHS sta in the roles and potentials o appropriately

constructed cost inormation.

Develop the evidence base on the potential and eectiveness o SLR and PLICS.

Executive summary

The NHS aces major cost reductions in the coming years.Two new costing tools are currently being implementedin trusts. The Department o Health (DH) recommendsNational Health Service (NHS) Acute Hospital Trusts adoptPatient-Level Inormation and Costing Systems (PLICS).Monitor recommends Service Line Reporting (SLR) preerablysupported by patient level data. A recent CIMA surveyprovides evidence that more than 70% o responding NHStrusts have implemented PLICS and/or SLR.

Results rom case studies and consultations with clinicians,senior nance managers, management accountants, andsotware suppliers1 show that great eorts are being madeto initiate PLICS and/or SLR implementation in the NHS. Thenext step, which now needs to be taken, is to obtain greaterengagement rom clinicians. Engagement is crucial or goingbeyond a box ticking exercise and assuring the eectivenesso these tools. Securing clinical engagement depends ona shit in the analysis o cost behaviour underlying newcosting tools rom allocating costs in ever greater detailto an analysis o activity and its resource consumption, thusactively supporting the management o cost. I this shitrom allocating to managing is not achieved then PLICS andSLR may become a more expensive top-down cost exercise,with little added value in comparison to traditional costingtools.

Fieldwork undertaken in several acute trusts hasdemonstrated that eective analysis o activity and resourceconsumption is being developed as a part o PLICS. PLICScan produce cost inormation which is more meaningul andrelevant or clinicians. It thus enables clinicians to considerboth cost and quality o care when taking decisions. There

is a potential danger that arises i trusts rst implementSLR with a view to moving on later to PLICS. I this is thecase then PLICS may be carried out as a top-down analysis.The benets o improved decision making inormed by costinormation may not be realised.

This report aims to highlight potential benets and pitalls oSLR and PLICS. It seeks to make clear the reasons underlyingthe eectiveness o such tools. To demonstrate these issueswe ocus in particular on the example o a central piece ohospital activity, examining cost calculation models o theoperating theatre.

Introduction

The National Health Service (NHS) aces a major challengeto transorm itsel into a patient-centred organisation,while at the same time saving over 20 billion between2011 and 2014. Managing care more eciently andeectively is crucial or tackling this challenge. As such, moresophisticated systems or analysing the nancial aspects ohospital activity are essential.

Until recently reerence costs have been the major NHScosting tool. Since 1998 the DH has collected and published

unit cost inormation, known as reerence costs. Reerencecosts are designed or central control purposes, with the mainobjective o calculating a national tari. It is substantially anannual exercise o collecting data at the level o the trust.Reerence costing is based on a top-down cost calculationmethod. To make cost inormation more credible and helporganisations understand their business/organisation betterand to better inorm national cost collection and PbR tarisetting, the Department o Health (DH) now recommendsPatient-Level Inormation and Costing System (PLICS).

1

This report is part o an on-going research project on managing in NHS acute trusts. The report was primarily unded by the Chartered Institute oManagement Accountants (CIMA), Grant Re. R 226, in collaboration with the Health and Care Inrastructure Research and Innovation Centre (HaCIRIC) atImperial College, London. In this context we carried out case studies in selected trusts and held ve discussion orums with invited experts.


3/12

Costing in the National Health Service: rom reporting to managing | 2

PLICS calculates costs at the level o the patient episode,thus allowing the more meaningul linking o costs toclinical data. Monitor recommends Service Line Reporting(SLR), the main output o which is nancial statement

o protability by service line. I used eectively PLICSand SLR together promise various benets. By providinginormation at operational levels, these tools may contributeto improvements in economic transparency, and therebyenhance clinician involvement in management. Togetherthese open up possibilities or both greater eciency andeectiveness o health care delivery.

...and thats the situation that a ew o us are now nding,with PLICS weve got the inormation and we can then start torefect upon that and decide well, how can we do it better, buti you dont have that inormation to start with you cant make

it betterClinician A

The eective implementation o PLICS and SLR requires aundamental shit in the analysis o cost behaviour however.Rather than top down allocation, costs should be traced sothat they can be actively managed. Rather than conceivingo overheads as an inevitable and unmanageable burdenor all, overheads can be unpacked and their value or thetransormation process questioned through activity analysis.The central question o eective management is not atwhich level to analyse or to report given costs, rather how toapproach the analysis o cost behaviour. This report argues

that greater attention to the basic principles o ActivityBased Costing (ABC) oers potential or trusts currentlyinvesting in PLICS and SLR to gain greater benets rom theirsystems.

For ABC to be eective, it is critical to dene what weunderstand by an activity. The number o activities carriedout by individual organisational members is so great thatit is not easible rom an economic perspective to use adierent cost driver or each activity. Thereore, activities areaggregated. Take the example o a corporate nance managerin a trust carrying out such activities as taking telephonecalls, responding to e-mails, reading reports, dratingdocuments, meetings with clinical units, meetings withnance sta, etc. These various activities could be aggregatedin more general categories such as 1) external reporting tothe DH and 2) internal reporting or clinical units. Activityreers thereore to the work process o what people areactually doing.

In the context o reporting in the NHS it is important tonote that the word activity is more commonly used to reerto specic outputs, oten clinical procedures that actuallyinvolve a variety o dierent work processes. Currently

much eort is spent in allocating all costs into such clinicalactivities. Whilst this results in an overall gure or cost,this gure does not help to make visible, and thereoremanageable, the various activities (and thus resourceconsumption) involved.

The introduction o PbR in April 2004 changed undamentallyboth the way health care is unded, and the role o costingin the NHS. Since PbR, NHS trusts are unded or a majorityo patients through a set price per HRG, which is reviewedand re-set every year on the basis o the reerence costs. Thisnew way o unding has consequences or the role o cost

calculations. It becomes increasingly important that the tariaccurately refects cost. Beyond this however, trusts and inparticular clinical sta need tools to understand how costsimplied in the tari relate to their clinical activity and thework (physical activities) that they perorm (and thereorehow resources are spent) may impact both clinical andnancial outcomes.

Table 1: Brie summary o the nature o reerence costs,PLICS and SLR

Frameworks Target output

Reerencecosts

An average cost o HRG treatmentepisodes used as the basis or settingthe tari.

PLICS Costs (and other clinical andoperational data) o treatment ora specic patient. Such data mightuseully inorm both reerence costsand SLR.

SLR A statement o revenues and costso clinical activities (and other key

indicators in the orm o a scorecard)to monitor and manage perormanceat service line level.

In a recent CIMA survey more than 70% o responding trustsreported that they had introduced PLICS and/or SLR. The vastmajority o trusts carry out costing reviews quarterly or evenmonthly (as opposed to the traditional annual review) andseek to involve clinicians and managers at more operational


4/12


levels. This engagement is essential i PLICS and SLR are todo more than just reassure stakeholders or regulators thatsystems are in place. Assuring the eectiveness o PLICSand SLR does not come rom ticking a box that cost data is

reported, but rather that cost data constructively inormsclinical decision making: taking the step rom reporting tomanaging2.

The orm o cost analysis underlying past costing toolspotentially intereres with this objective o using costinormation or management. Hospital costing has been amajor concern or the government since the creation o theNHS in 1948 (MoH, 1948). Departmental costing, specialtycosting, management budgets, resource management, HRGand PbR are some o the hospital accounting innovationswhich were implemented over the decades. The design o

these costing tools was mostly driven by central controlpurposes.

An important actor in the limited engagement o clinicianswith these tools is the underlying approach to the analysiso cost behaviour: the allocation o overheads and indirectcosts onto clinical units or activities. Traditional costingmodels start rom the general ledger and allocate overheadsacross a growing number o HRGs. The calculations areso complicated and unclearly linked to clinically relevantcategories, that clinicians cannot gain an understanding othe cause and eect relationships between their decisions,resource consumption, and clinical results. The output is

not inormation that can be easily used by ront line stato improve perormance and drive value or money. Thisapproach to the analysis o cost behaviour may point outproblems, such as costs o a certain HRG being abovereerence costs, but does not provide any guidance on howto x the problem.

This orm o analysis o cost behaviour potentially infuenceshow PLICS and SLR are developed and currently used.Complex allocation models or overheads and indirect costsare in many cases still the underlying method o producingPLICS and SLR data. Higher granularity o data is achieved(patient or service line level), requiring more resources to

produce. I the top-down approach is still adopted thereis little advancement in making data more transparentand usable. As clinicians cannot easily relate this cost datato their physical activities, the drawing o operationalconclusions is greatly impaired. Thereore, the danger existsthat PLICS and SLR are reduced to an expensive exercise inproducing and reporting numbers, with little infuence inclinical decision making.

In our view the key question is not about granularity andlevel o analysis, but rather the approach (see table 2). A

more detailed level o analysis (e.g. service line or patientlevel) is not the driver or eciency and eectiveness initsel. Rather, what is needed is an ABC approach to coststhat does not regard overheads or costs in general as

inevitable. Instead costs and their contribution to healthcare are questioned through activity analysis and potentiallytransormed in direct costs.

Activity analysis is designed to establish meaningulrelationships between consumed resources and activity. Costspreviously classied as overheads may be translated intodirect costs out o general ledger categories. In this way theyare not allocated, but traced. I the logic o tracing a xedcashfow cost (say clinician time to patients through minutesspent) is accepted, then in principle the logic o tracingother xed cash fow costs is equally possible given the right

cost pool structures and cost drivers. ABC urther suggestsremoving fuctuations in reported cost due to changesin activity levels through the use o capacity cost driversand the separate reporting o unused capacity. I an ABCapproach is lacking, PLICS or SLR/SLM will add little valueto cost management in comparison to the long establishedindicators at clinical level length o stay or cost o drugs.

Table 2: Summary o dierences in approach to analysingcost behaviour

Approach Summary

Top-downull cost

Costs are tagged as direct, indirect,overhead (per DH guidelines). A complextop-down process cascades allocatedoverhead until the ull cost is collected atthe level o HRGs. Danger that overheadsare regarded as xed or inevitable. Theirburden must simply be shared, sinceoverheads are, by denition, not logicallytransparent in their behaviour.

ActivityBased

Costing(ABC)

Costs pools are translated rom generalledger categories to a structure refecting

dierent activities making coststransparent and manageable (viewedas more important than an attempt toproduce a notional ull cost). This oersgreater potential to evaluate eectivenessand eciency in terms o their valuecontribution to the overall businessprocess, patient care. Can be moreexpensive, so judgement is necessary asto how ar to take this.

2 Hence the increasing emphasis on Service Line Management (SLM)


5/12


From reporting to managing

Even though there have been major improvements in thedata quality o costs and taris, many clinicians do not

recognise these improvements:

I think thats been one o the big problems in the NHS: wehavent had the inormation, weve got no idea what anythingcosts.

Clinician B

Weve been doing costing or 15 years, I dont know,whenever, when we rst started looking at it, at what thingscost, and here we still are sitting saying, Actually, nobodyknows what things cost, theres a tari that nobody agreeswith, youd think that over all those years they could have gotit a bit nearer to being right.

Clinician C

This perceived lack o improvement in the quality o costinormation is linked to the prevailing approach to theanalysis o cost behavior. Allocating the bulk o overheadsand indirect costs in a topdown manner onto dierentclinical outputs may satisy the objective o producing atari, but it is not satisactory or inorming clinical decisionmaking.

I costing systems are implemented ollowing a top-downmethodology o allocating costs, the resulting costinormation gives clinicians little added value in comparisonto traditional indicators such as length o stay. In consideringully absorbed ward costs the only aspect o operationsclinicians may understandably control is volume. The detailedanalysis o overhead spending in relation to their activitiesand choices remains beyond their reach and understanding.Given this, a precise gure o overheads oers them nourther benets over a general understanding that reducinglength o stay is benecial (spreading the overhead morethinly across more patients). In the ollowing a clinician isquestioned on his uture use o PLICS:

No because I think length o stay indirectly tells you what you

want to know. PLICS exists as a nancial package but the waythe cost base has been built up, its not suciently accurateto give data that I would say is wholly reliable at the moment.[PLICS] is supercial, generalised, lacking real sensitivity andwe could do so much better.

Clinician D

The danger then is that PLICS and SLR are not used ormanaging, but only or box ticking to reassure externalregulators and other stakeholders:

Yeah, and Service Line Reporting were doing quarterly, butit does go out because its an important box to tick or [theregulator] , [] , but other than a ew people spotting thattheyve got an overhead it hasnt really changed anything.

Senior manager A with responsibility or PLICS

In the interviews and the discussion orums there was aconsensus that the development o PLICS guidance andstandards initiated by the DH were a oundation or a newapproach to cost management in the NHS.

I mean, weve got this PLICS guidance, costing guidance,which these guys have been involved in, and thats a reallygood start and thats the kind o stu you need to build on.

Senior manager B with responsibility or PLICS

But while the guidelines played a major role in initiating andsupporting the implementation o PLICS, they now have tobe developed and rened. A central area or more detaileddiscussion is the treatment o overheads and indirect costs.In particular a more detailed explanation o an ABC approachand the importance o building up cost data around actualphysical activities i.e. work that people do (and not simplyoutputs, or departments) can make a wider range o costsamenable to management. This emphasis on physical activityas the appropriate ocus or cost analysis should be ar moreclearly explained in such guidance, i the hoped or benetso investments in new costing systems are to be achieved.

In this context it is important to dene more clearly in theguidelines what is understood by an activity in an ABC sense.Otherwise there is a danger that even or things like cliniciancosts, which should in principle be direct costs, simpledesign errors (such as grouping clinicians with very dierentpatterns o activity) mean that the costing system may stillsystematically misrepresent them even with an appropriatecost driver. In the next two sections we discuss how a changein the analysis o cost behaviour produces cost data, which isuseul or operational management, in particularly in regardsto clinical decision making.

The case o the operating theatre (romdepartment to activities)

By analysing the dierent cost calculation models or theoperating theatre used by various NHS trusts we can showhow these dierences infuence the degree to which costdata is useul or management. Many trusts included inthis research started with a simple cost calculation model,consisting o a single cost pool that assembled all costsassigned to the operating theatres. The cost driver is thenminutes spent in theatre. To calculate costs per patient, the


6/12


total costs are then divided by the total theatre minutes. Thisresults in a cost per minute. Based on the operating register,recording the time each patient spent in the operatingtheatre, the costs per specic patient can be calculated.

Figure 1: Very simple model

Even though this model allows the calculation o costs atthe patient level, it gives only a very approximate cost orany individual patient. The main problem is the very loosecause eect relationship between clinical activity and theircosts. The operating theatre is conceived as an administrativeunit to which a certain amount o costs are assigned. Thesecosts are then treated as overhead costs and are allocatedrather than more logically traced onto the patient episode.Whilst minutes do infuence costs, they are not the onlyrelevant actor. Another signicant driver is the number osta carrying out the surgery. Thus while the number osurgical sta may vary greatly between dierent procedures

in practice, the use o such a simple model results in theapplication o the same cost per minute.

In one trust that we studied, this simple model was rened todistinguish between dierent specialties, or example cardiacsurgery, orthopedic surgery or plastic surgery. Costs or staworking within dierent specialties are captured in specialtybased cost centres. Specialty based operating theatres easilyallow or identiying other running costs per specialty. Thiscost calculation model thereore developed a cost pool or

each specialty. The activity drivers are still the minutes spentin each theatre. The results however show the dierent costper minute across the dierent specialties.

This more rened cost calculation model produces costdata useul or management i the resources spent withina specialty do not vary rom one procedure to another, orexample the number o sta carrying out all cardiac surgery.But or many specialties, the resources spent in carrying outsurgery may vary greatly, particularly in regards to numbero sta. As a result, both the simple and intermediate costmodels share the same problem in that costs are linked witha unit rather than actual activities. This does not allow or theestablishment o meaningul links between physical activityand resources consumed. I the bulk o clinical costs aretreated as overhead cost and are allocated onto the patient

episode on the basis o the same costs per minute or eachspeciality, the model cannot take into account the variety oresources spent. It eectively applies the same average costper minute to a procedure, regardless o whether it involvedve clinicians or one. It does not recognise that some costsare not driven by time but by, or example, the number ooperations perormed.

Learning rom cost data requires a clear signal that certainactions results in specic cost outcomes. The logic ordirect costs is clear. I less time is spent with a patient thenreported cost goes down. This saving must o course beconsidered in relation to treatment quality (which is why

PLICS combines clinical and cost data). When it comes tooverheads however the above cost calculation models do notproduce reliably clear eedback. In such systems, overheadcosts, which by denition do not relate to patient volume,are being treated as though they did. Whilst the clinicianmay spend less time per patient and consequently see morepatients, because they receive overheads rom a generalpool whose allocation base sits across multiple clinicians andprocedures, the quantum o overhead that they receive per

Costs/per theatre minute

Cost driver: total number

of theatre minutes

Cost pool with all

costs for all theatres

Costs/per orthopedic

surgery minute

Cost driver: totalnumber of orthopedicsurgery minutes

Cost driver: totalnumber of cardiacsurgery minutes

Cost driver: totalnumber of eyesurgery minutes

Cost pool orthopedic

surgery

Costs/per cardiac

surgery minute

Cost pool cardiac

surgery

Costs/per eye surgery minute

Cost pool eye

surgery

Figure 2: Intermediate model


7/12


patient may in act rise. Overheads can thus be regarded asunsuitable or operational control since the eedback theygive is noisy at best.

A second trust that we studied rened their cost calculationmodel to deal with this problem by engaging in activityanalysis and cost tracing. In the rst step, the trust identiedthe dierent activities that are carried out in the operatingtheatre. Secondly, the trust calculated costs or the dierentactivities. This was done by establishing meaningulrelationships between costs and activity. Instead o allocatingcosts onto just clinical activity, cause-eect relationshipsbetween resources consumed and physical activities wereestablished. The relevant unit o analysis or such activities inthe operating theatre is the surgery session. A session is thenumber o hours booked in theatre by a surgeon, or example,

a our hour block.The system captures the elements that the drive costs o asession. Through activity analysis the ollowing activities andcost drivers were identied or a surgery session:

preparing the operating theatre beore theatre use

perorming anesthesia (anesthetic sta and anestheticdrugs)

perorming the surgery (sta)

consumable items during surgery

recovery o the patient

cleaning up ater theatre use overheads: management administration, training, catering

and others.

Moving rom the operating theatre as a department, in thismodel the operating theatre is seen as the site o a variety ophysical activities. Once these are identied, the next step inthe ABC methodology is to understand what drives the costincurred in carrying out these activities. In approaching this

task it is important to always apply the principles ocost/benet analysis in terms o how much detail andrealism to aim or. The question to ask is does the eort totransorm a particular block o overheads into direct costs

through an analysis o activity pay back in terms o decisionmaking benets obtained rom such eorts?

In the trust that we studied, the analysis o cost drivers, andthe tradeos between detail and managerial useulness aresummarised in table 3. Clearly there are many simpliyingassumptions, and in each case the simpliying assumptionsrefect both the cost o capturing more precise data, as wellas the potentially limited learning that might arise romdoing so.

The ollowing renements were taken into account as themodel was urther developed:

i surgery takes place within normal working hours or outo hours at weekends, evenings

a charge or early starts and over runs

an availability charge or the priority tables out o hours.

Table 3, on page 7 (overlea), shows how or each activitya relationship between the activity and the consumedresources is established ollowing the cost /benet principle.In some cases this leads to charging a standard rate, e.g. orpreparing the theatre. In other cases the consumed resourcesare related to an activity on the basis o a cost driver, such

as time. Anaesthetic drugs, or example are charged on thebasis o the length o time the patient is anaesthetised. Thedata o the time anaesthetised is available in the system andcan be retrieved easily. Costs or nurses take into account thenumber o nurses present, which can also be captured in thesystem. However as it is too costly to include the actual stacost or each nurse, a charge rate is applied or each nurse.For senior clinicians the actual sta cost is assigned to thesession, while or juniors it is again a charged rate.

Different cost drivers for each activity

Cost of activity one Cost of activity two Cost of activity n

Patient cost = Activity1*Driver1 + Activity2*Driver2 + ... + An*Dn

Figure 3: An example of a more complex cost equation


8/12


A strong advantage o the data produced by the model ingure 3 is that it draws rom activities that are clinicallyrelevant.

I think the advantage o [PLICS] is that we drill down romthe surace, down to the patient, and you can show the costs,my costs saying This patient, [Name], was in or this periodo time, was in theatre ve or three and a hal hours, he hadthese pathology tests, he started o rom [X] Ward and heended in [Y] Ward, and the clinician goes Yeah, I believe that,thats what happened, and then you move the discussion

away rom I dont believe the numbers to What are we goingto do about the numbers?..

Senior manager C With responsibility or PLICS

Providing such inormation invites clinicians to engage inboth personal and collective refection on the eciencyand eectiveness o their activity. PLICS ollowing an ABCapproach produces the evidence base or comparisonsbetween clinical practices and can become a tool orinorming health care practice, the redesign o practice andthe denition o patient pathways.

Table 3: Summary o activities and cost drivers

Activity Appropriate cost driver

Preparing the operatingtheatre beore theatre use

A standard charge is applied. This assumes that the length o time o preparing the theatredoes not vary greatly between patients or procedures.

Activity o anesthetic nurse A standard charge or providing anaesthetics is applied or anesthetic nurse, distinguishingwhether it is a general or local anesthetic. This assumes that the sta costs o perorminggeneral or local anesthetics do not vary greatly between patients.

Activity o anestheticphysician

A standard charge or providing anaesthetics is applied or anesthetic physician,distinguishing whether it is a general or local anesthetic. This assumes that the sta costso perorming general or local anaesthetics do not vary greatly between patients.

Anaesthetic drugs Charge or anaesthetic drugs based on a weighting o the length o time the patientis anaesthetised. The time taken into account is rom entering the anaesthetic roomuntil entering the recovery room. This assumes that consumption o these drugs is wellcorrelated with time.

Activity o the nurse orperorming the surgery

Costs or the surgical nurse are charged on the basis o the session schedule. The timetaken into account is the knie to skin until closure. The model charges a rate accordingto the number o sta present. It does not take into account actual sta costs (e.g.dierentiate between agency and ull time sta on the day).

Activity o the physician orperorming the surgery

Costs or clinicians are charged according to the clinician planning schedule oprogrammed activities. The time taken into account is knie to skin until closure. Thesystem takes into account actual sta cost or seniors with a standard rate or juniors.

Consumable items during

surgery

Currently this is done on the basis o itemised lists o consumables by procedure. This

system will be replaced in the near uture by a new bar coding device capturing the actualconsumption o consumables or each patient.

Cleaning up ater theatre use A standard charge is applied. This assumes that the material used and the length o time ocleaning up the theatre does not vary greatly between patients or procedures.

Recovery o the patient An average charge or recovery or each patient. This does not take into account patientdependency and length o time in recovery. A dierent charge is applied to those patientsentering overnight intensive recovery.

Overhead activity A charge is applied or management administration, training, catering and others on thebasis o the charged time.


9/12


I think you could use [PLICS] as a catalyst or nationalcomparisons, to encourage people to change practice. I theycan see examples o good practice elsewhere. Thats alwaysbeen part o medicine, you know, when I was a lad as a junior

surgeon a long time ago, people used to be kept in ten daysater a hernia operation, now its a day care, is the examplethats always quoted. So I mean Im sure there are things wecan do

Clinician E

The purpose o ABC is to produce usable data, and animportant aspect that supports usability is the separateconsideration o the cost o capacity. In the simple systemmentioned, various costs are traced to patients according tothe number o minutes actually spent carrying out physicalactivities with them. The cost per minute traces this to the

total minutes actually used. This means that in a sessionwhere the ull our hours are used, cost per minute will belower, than in a session where only two hours are actuallyused or physical activities. This correctly gets across theintuition that it is costly to not use the ull session. Howeverthe higher cost message needs unpacking to understand thatit is utilisation and not cost o resources (e.g. more costlyclinical sta) that have caused the cost rise.

Calculating a cost o unused capacity separately has twoeects that improve the usability o the data. Firstly it cutsout fuctuations in cost per minute o physical activity thatis caused by changing capacity usage. Secondly, by reporting

unused capacity separately, it raises the issue or discussion.The implications o unused capacity may vary. In somecases, it may be deemed clinically necessary, and thereoremust be taken into the baseline consideration o resourcesrequired or certain procedures. In other cases, it may be thatrearrangements to the location or scheduling o proceduresmay oer more ecient ways to work. An extension othe capacity analysis would be not only to perorm suchan analysis or the operating theatre as a whole, but perphysical activity. This would give a much more renedunderstanding o how capacity is used at operational levels.

Managing costs is not a problem o having a perectlydesigned system. In act there is no perect system thatcan apprehend all the complexity involved in clinicaldecision making. The cost o attempting such a systemwould signicantly outweigh the decision benets it mightyield. The point o a system such as PLICS based on ABC isthat it provides clear signals o issues to discuss as well asproviding relevant inormation to inorm decisions as to howto undertake activity. Such questions may be consideredin other orums, however i investments are to be made inPLICS it seems that enabling it to support such a role wouldbe prudent.

The case o the accounting and nance unction rom department to activity

Changing the approach to the analysis o cost behaviour

is not only key or managing clinical activities, but also ormanaging the activities o the central administration. I costscurrently labelled as overheads are accepted as inevitablethen there is a risk that they are orgotten and excludedrom perormance analysis. ABC, by questioning the notiono overheads or indirect costs, seeks to make a greaterrange o cost amenable to direct management. The dangero top-down models is that overheads cannot be easilyunderstood, nor their benets eectively considered. I this isthe case, then the danger is that they may rise or the wrongreasons.

None o the nancial managers taking part in this researchroutinely reviewed inormation about the percentage ooverheads as a component o overall costs, nor the trend(o growth, or decline) in recent years. ABC advocatesthat overheads are not a act o nature, but a act o costpool structure. Overheads can be transormed into directcosts through activity analysis. In the case o the nancedepartment, costs in the overhead cost block are typicallyshared out among clinical specialities. Having such costsimposed on them does not enable clinicians to thinkabout this activity in relation to what they are doing, norappreciate the benets such overheads might oer insupporting them. While some o the activities in a nance

department may indeed concern all specialities equally, thereare various activities which are directly related to certainHRGs. One example is the preparation o cost analysis ornegotiations with Primary Care Trusts (PCTs), or example.

For our patients on the costing system, we were able todemonstrate that we were bringing in women to have amastectomy, and then some months later theyd come backand have a reconstruction under the tari, and get paid 6,000or that procedure . Our general surgeons and plastic surgeonswanted to work together in the theatre and do it all in one day[] and you only get paid once under the tari, so it costs you12,000 but you only get 6.000, but because o PLICS we

could demonstrate that to the PCT, and theyve agreed to takeit out o tari and pay us separately or it.

Management accountant A


10/12


An activity analysis may allow not only more transparentand logical tracing o overhead costs, but also supportthinking around the redesign o the overall activity fow.For example activity analysis may reveal redundant or

overlapped activities between central and operational units.In such cases, the sta o central and clinical units mayrefect jointly on how to improve the fow o inormation atcritical interaces.

One such interace that was requently highlighted in ourresearch was between clinicians, coders, managers and PCTs.A monthly reporting cycle can contribute to a more ecientfow o activities at this interace.

The tari thats charged is provisional or the rst [] isit 30 days? And then the other 30 days, thats it, you cantchange it, in other words, i you can go back to the PCT, the

commissioners, within that period and say actually no, wevemade a mistake on this coding here, it wasnt this, it was that,we now want more money then theyre obliged to pay it.I you go past that 30 days, then there is an obligation, so iwhat this is doing is giving you greater insights into accuracy,youve got coding and so on, and it does. So sometimes itjust completely makes out that youre making losses becauseyouve cocked up the coding, then you can go back to thePCT and ask or the right amount. So i youre doing it [PLICS]on a quarterly basis, then you can only do it or a very limitedperiod.

Management accountant B

Another critical interace emerging in the discussion orumswas between PLICS and SLR and other nancial managementprocesses. Participants pointed out that there is oten littleor no connection between PLICS and SLR on the one handand management reporting or budgetary processes on theother. Potential conficts between these may then become apotential source o ineciency and ineectiveness.

Im interested in PLICS and Im interested in providing thoseoutputs, but it seems intuitively to me, instinctively, it seems iwere going to all this bother with PLICS and producing theseoutputs and the robustness and weve had our systems set up

with people lined up, then why dont we do our planning romthat?Why dont we do our planning and align it? one othe by-products [] o PLICS output would be to service theplanning o it and all the decisions that go along with that. So at the moment most organisations, theyre totally divorcedreally, []. In our organisation all the divisional structureis totally tied into the budget, the budget, the budget, thebudget, operation, the divisional manager, the nancemanager, clinical nance managers. Lately weve started tolook at service line reporting, but they dont really get it andits like, you know, as someone was saying earlier on, why

arent people looking at service line reports, why not? One othe reasons, possibly, is we need to line it up, we need to doour planning rom PLICS.

Senior manager D with responsibility or PLICS

A urther critical interace or the nance department is thelink between external and internal reporting tools. Managersin the nance department argued that the dominating roleo external reporting, the ledger, intereres with using PLICSand SLR or management purposes. With the main objectiveo external reporting being to produce numbers or externalevaluation, the philosophy o using such numbers or internalmanagement is unclear. The dominating role o externalreporting refects the stang o many nance departmentshowever.

To explore this issue in more detail we carried out a smallsurvey with the Network o Foundation Trusts. Eightquestions about the sta resources spent on PLICS and/or SLR and the stang o the nance unction were sentto all member trusts. 27 responded, o which all hadimplemented or were implementing PLICS and/or SLR. Acrossthe responding trusts there was an average o less than 1.5ull-time equivalents (FTE) working on PLICS and/or SLRper trust out o an average o 55 FTE working in the overallaccounting and nance department. Even in the largesttrusts with more than 100 FTE working in accounting andnance the average number o FTE working on PLICS and/orSLR was two.

Given the potential signicance o PLICS data, this suggeststhat relatively ew resources are spent on linking numberswith clinical activities to actually manage activities. Hencear more detailed activity analysis o the nance unction(or example) is not a peripheral issue to the managemento NHS trusts. The organisation o the nance unction playsan essential role in shaping the eectiveness o wider clinicaland operational management. Again, ollowing rom ABC, theoverheads o the nance unction (and by extension otherorms o overhead) must be open to question and discussionto help ensure that overall resources are spent eectively andeciently to support patient care.

Conclusion

Our research questions the dominant approach to theanalysis o cost currently ound in the NHS that assumesmany costs are inevitable. The emphasis on cost recoverythat began with the tari, no longer serves the NHS well inan era that increasingly emphasises the need or eciencyand eectiveness. ABC was the inspiration or PLICS, and soin approaching PLICS and SLR we argue or greater attention


11/12


to unpacking and discussing currently opaque cost structures.The vocabulary o cost analysis used in the NHS dividingcosts in overheads, indirect and direct costs contributeslittle to the eective management o cost, activities and

treatments. Rather than ollowing a logical denition o cost,based on cost behaviour, overheads are most requentlyexplained with reerence to their likely organisational/departmental source. Typical examples emphasise that theyare something ar away rom the patient, contributing toa notion o their unmanageability. Closer to the patient,direct costs are also requently dened by location and notbehaviour. Here again the danger is that cost manageabilitymay suer. Given that the NHS must make major costsavings in coming years, this report argues that ABC analysiso cost behaviour is a tool that oers many attractions,outweighing the cost o undertaking it.

The transparency generated through activity analysisstems not rom allocating costs more accurately, butrom questioning the value o activities or the coretransormation processes o the organisation. An activityanalysis starts by identiying and mapping out activities.

Identiying activities and recognising how they are linkedwith patient care is an important step to transparency andan evidence based approach. Such an analysis oten revealsineciencies in terms o quality, time and cost which can

then be the basis or the redesign o activities in order tocreate a more eective and ecient activity fow, ultimatelyimproving patient care.

In contrast to traditional costing approaches which can belargely produced (and consumed) in the central accountingand nance department, ABC relies on the participationo all those involved in the transormation processes.This inormation is more relevant or others since itsundamental unit o analysis is actual activity (as opposedto abstract constructions o the general ledger, or HRG, orexample). Only i costing is signicant to all those involved

in the transormation processes might there be a shit inemphasis rom reporting cost to managing perormance.Inormed by the principles o ABC PLICS and SLR oer greatpotential to position clinicians where they belong, as theessential custodians o the delivery o ecient and eectivepatient care.

Acknowledgements

The research was nancially supported by HaCIRIC. We thank the ollowing contributors or their time, the use o data,inormation, case study material or eedback, guidance and support: participants o ve discussion orums on ABC, PLICS and

SLR including senior PLICS managers, management accountants,nance managers, clinicians and sotware suppliers; cliniciansand managers who were object o the case studies; The Network o Foundation Trusts; those who commented the drat andthree anonymous reviewers.

Reerences

CIMA, 2009, A picture o health Service cost data in perormance management in the NHS, October 2009.

Department o Health, 2009/10, NHS Costing Manual: Gateway reerence: 13659.

Department o Health, 2009/10,Acute Health Clinical Costing Standards: http://webarchive.nationalarchives.gov.uk/+/www.dh.gov.uk/en/Managingyourorganisation/Financeandplanning/NHScostingmanual/DH_080056

Ministry o Health, 1948, Statutory Instrument1414. London: HMSO.

Glossary

ABC Activity Based Costing NHS National Health Service

CIMA Chartered Institute o Management Accountants PbR Payment By Results

DH Department o Health PLICS Patient-Level Inormation and Costing System

HRG Healthcare Resource Groups SLR Service Line Reporting

MoH Ministry o Health HaCIRIC Health and Care Inrastructure Research andInnovation Centre


12/12

ISBN 978-1-85971-670-0 (PDF)Chartered Institute of

Management Accountants

26 Chapter StreetLondon SW1P 4NP

United KingdomT. +44 (0)20 8849 2285E. [email protected]

July 2010, Chartered Institute o Management Accountants

Documents

Costing in the NHS