Mark H. Lukasik, MD, CCFP Glenn Pratt, MD, FCFP

The Telephone:An Overlooked Technology

for Prevention in Family Medicine

SUMMARYAnnual influenza vaccination has long beenrecommended for the elderly population.Despite this recommendation, immunizationrates have remained very low. This studymeasured the effects of two approaches to theprovision of influenza immunization to the65-years-and-over age group in a single familypractice. The "drop-in" group (N=123) wasinformed of the availability of the vaccine atvisits made during the vaccination period. The"phone" group (N= 120) was notified of theavailability of the vaccine by telephone andwas invited to come in for the shot. Animmunization rate of 50.8% for the "phone"group and 26.8% for the "drop-in" group was

obtained (P=.0002). These results contraststrongly with the overall immunization ratesof 5.9% and 9.5% obtained during theprevious two years, when no activeimmunization policy was in place. Thetelephone approach was found to benefit thetype of patient at greatest risk from influenza:the chronically ill and the aged. It is dear thathaving a defined immunization policysubstantially improves the provision ofinfluenza vaccination. The authors discuss theeffectiveness and practicality of theseapproaches to the delivery of influenzavaccine and their applicability to other formsof prevention in family medicine. (Can FamPhysician 1987; 33:1997-2001.)

RESUMELe vaccin annuel contre la grippe est depuislongtemps recommande pour les personnes agees.Malgre cette recommandation, les tauxd'immunisation sont demeures tres bas. Cette etudemesure les effets de deux approches a la vaccinationcontre la grippe chez les gens de 65 ans et plus dansun contexte de pratique familiale. Ceux qui se

presentaient sans rendez-vous prealable (n= 123)etaient informes de la disponibilite du vaccin lors devisites faites pendant la periode de vaccination. Unautre groupe que nous avons rejoint par telephone(n= 120) fut informe de la disponibilite du vaccin etinvit6 a se presenter pour y recevoir le vaccin. Nousavons obtenu un taux d'immunisation de 50.8%pour le groupe rejoint par telephone et de 26.8%pour celui << sans rendez-vous >> (P=.0002). Lesresultat, contrastent fortement avec les taux globauxd'immunisation de 5.9% et de 9.5% obtenus au

cours des deux annees prece;dentes alors qu'iln'existait aucune politique active d'immunisation.L'approche par telephone s'est averee avantageusepour les patients a plus haut risque de contracter lagrippe: les malades chroniques et les personnesagees. Il est evident que l'elaboration d'une politiqued'immunisation claire ameliore substantiellement letaux de,vaccination contre la grippe. Les auteursdiscutent l'efficacite et le caractere pratique de ces

approches a la vaccination contre la grippe et leurapplicabilite aux autres formes de prevention.

Key words: elderly, influenza vaccine, preventive medicine, telephone

Dr. Lukasik is a former residentin Family Medicine at theUniversity of Western Ontario. Dr.Pratt is Professor in theDepartment of Family Medicine atthe University of Western Ontario.Requests for reprints to: Dr. M.H.Lukasik, 1-497 Beechwood Drive,Waterloo, Ont. N2T 1H8

A NNUAL INFLUENZA vaccina-tion in the elderly has been re-

commended by public health and im-munization agencies in NorthAmerica for over two decades.", 2 Itsannual use is advocated by the HealthMaintenance Guide, formulated bythe College of Family Physicians ofCanada in 1983.3 Numerous studieshave demonstrated the effectiveness

of the vaccine in reducing morbidityand mortality in high-risk groups, in-cluding the elderly, especially duringepidemics.4-8 In spite of all this atten-tion, it is believed that less than one-fifth of the elderly population in Can-ada and the United States receivesinfluenza vaccination each year. 1, 9-11

In the past, concerted attempts havebeen made to increase the influenza

CAN. FAM. PHYSICIAN Vol. 33: SEPTEMBER 1987

INNIMmosionsommamoffmmmmmm I 01009--, mom

in 1. .Sam 0 1INNOWMENIII.I.... men=

1 997

immunization rate. In response to apredicted "swine" influenza epi-demic in 1976, a mass influenza-im-munization campaign was conductedin the United States. As a result influ-enza vaccine delivery rates rose to40% in high-risk groups (the elderlyand the chronically ill) and to 25% inthe general population.2 12 "Re-minder Postcard" studies have in-creased immunization rates in certainlocal areas. In one successful study a59.7% vaccination rate was recordedfor patients who had received thepostcard as compared to a rate of 30%for those who had not.'3 Other post-card studies have been much less suc-cessful. 12' 14 Recent Canadian studieshave investigated the effect of post-cards and the telephone reminders onthe influenza immunization rate.Frank and his colleagues obtained arate of 43% after using reminder post-cards. This rate rose to 55% when afollow-up telephone call was made tonon-responders.'5 McDowell and hiscolleagues (1986) obtained rates of35.1% and 37% for patients remindedby postcard and telephone respec-tively, compared to a rate of 9.8% forthe control group.16 The long-term ef-fects of these influenza-vaccine pro-motions is unclear.The main objective of this study

was to compare two methods of rais-ing the influenza immunization rate ofthe elderly population of the practice.To achieve this objective the follow-ing approaches to influenza vaccinedelivery were undertaken.The telephone is an inexpensive but

overlooked technology that can beused to enhance immunization. Toassess its effect, patients in the phonegroup were contacted by telephone asan initial outreach measure.To assess the impact of simple of-

fice policies, the drop-in group wasformed. In this group no outreachmeasure was undertaken, but the in-ternal office environment was used topromote influenza vaccination: re-minder stickers were affixed to patientcharts, a bulletin was posted in thewaiting room, and the advantage ofinfluenza vaccine was routinely dis-cussed with the elderly during theirvisits to the Family Medical Centre.The physicians involved in the

study investigated the effectivenessand practicality of these approaches tothe provision of influenza vaccine.They also examined the characteris-

tics of those patients in each groupwho received the vaccine.

MethodThe study was conducted from

mid-September to December 1985, ina single family practice at the VictoriaFamily Medical Centre (FMC), ateaching practice affiliated with theUniversity of Western Ontario, inLondon, Ontario. The "Health CareTeam" involved in the study con-sisted of the staff physician, two phy-sicians in residency training, a regis-tered nurse, and registered nursingassistant.

The study population included allactive registered patients in the prac-tice, 65 years or older. Patients chron-ically hospitalized or in nursinghomes were excluded from the study,as were those who were unable tocommunicate by telephone or whowere housebound.

During previous years no outreachmeasures had been taken in the prac-tice nor had there been any promotionof internal influenza vaccine. Whilethe staff physician acknowledged thevalue of influenza vaccination for theelderly, no specific policies were inplace. Prior to the study, patientswere vaccinated against influenza ifthey specifically asked for the proce-dure, or if the physician believed it tobe particularly valuable.The study population was divided

into two groups, the "Drop-In"group and the "Phone" group. Aftera random start patients were alter-nately assigned to each group, thoughrelated patients and those living in asingle household were kept in thesame group.A brightly coloured sticker was ap-

plied to the charts of the entire studypopulation as a reminder to thehealth-care team that the study wasunder way and that they were ex-pected to promote the flu vaccine.

All collaborators in the study metand agreed on a similar approach. Thepatients would be told, whether bytelephone or in the office, that thevaccine was available, and that theywould be given a shot if they wished.Moreover the collaborating physiciansagreed on responses to be given toquestions about side-effects and thevalue of immunization.

Patients in both study groups whovisited the FMC during the immuniza-

tion period were treated the sameway. On entering the examiningroom, the patients were informed by anurse or a nursing assistant that thevaccination was available. At the endof the visit, the physician addressedany unresolved questions about theimmunization, and the shot was givenif the patient wished. Flu shots werenot given to patients with an allergyto eggs or to patients who had reactedto previous influenza immunization;they were postponed for those patientssuffering an acute febrile illness. Asingle handmade 8" x 11" advertise-ment, bearing the caption "Be KeenAbout Flu Vaccine", was posted inthe waiting room.

The only methodological differenceshown the two groups was that somepatients in the phone group received atelephone call. If a patient visited theFMC before the staff phone call wasmade, no call was deemed necessary.The telephone calls were made in ap-proximately equal numbers by thestaff physician, a registered nurse,and a registered nursing assistant; theduty of phoning was assigned in turn,systematically, with random start.Each telephone caller had been asso-ciated with the practice for at leastfive years. The structural design ofthe study is illustrated in Figure 1.

Each week the phone callers wereprovided with a call sheet of fivenames. If the caller made the assignedcalls before the week was up andasked for another list, it was pro-vided. The calls were made during of-fice hours. Patients were told of theavailability of the influenza vaccineand informed that they could receiveit during a regular visit ("appoint-ment") or schedule a time with thenurse ("clinic"). If the phone was notanswered after seven rings, the callwas considered a failed attempt. Abusy signal was documented, but wasnot considered an attempt. A maxi-mum of three phone attempts weremade to each household, and the at-tempts were always spaced at least ahalf day apart. No further follow-upphone calls were made.

Following the immunizationperiod, the collaborators reviewed allcharts. The analysis was done withpatients in their originally assignedgroup, whether or not they had re-ceived a phone call. Thus the projectwas an "intention-to-treat" analysis,the type which has been generallv

acknowledged as the most appropriatefor randomized trials. 17

DefinitionsFor the purpose of the study the

following definitions were standar-dized. A "chronic illness" was de-fined as a disorder of more than threemonths duration of the cardiovascu-lar, pulmonary and/or renal systems;metabolic disease; severe anemia;and/or compromised immune func-tion. Patients subject to chronic ill-ness are thought to be at moderate tohigh risk of serious illness from influ-enza as compared to the general popu-lation." 18, 19 Hypertension was in-cluded in this category if the patientrequired anti-hypertensive medicationfor control of the condition. A partici-pating physician determined thenumber of chronic illnesses fromwhich a patient suffered, whethernone, one, or two or more.An "adverse reaction" was defined

as a reaction to either a medication ora vaccination (of any kind) as docu-mented on the patient chart. "Pre-vious influenza vaccination" was de-fined as influenza vaccination givenin 1983 or 1984. For documentationof "years attending the centre" pa-tients were grouped into categories offewer than 15 or 15+ years. Underthe notation "household composi-tion" patients were classified as beingalone, or living with a spouse or other

ResultsThe two study groups were com-

pared in relation to a number of de-Figure 1

mographic and health factors: sex,mean age, marital status, householdcomposition, mean number of visitsto the centre, years attending theCentre, adverse reactions to medica-tion, the presence of a chronic illness,and the number of chronic illnesses.The analysis did not show a signifi-cant difference between the studygroups for any of these factors.

Table 1 illustrates an overall influ-enza immunization rate of 38.7% forthe entire study population. This fig-ure represents a seven-fold increaseover 1984 levels. It is, however, thecombined rate of the two studygroups. A 50.8% immunization ratewas recorded for the Phone groupcompared to a rate of 26.8% for theDrop-In group. The influenza immun-ization rates of the Phone group andthe Drop-In group were increasedover the previous year by 43.2% and22.7% respectively'(P=.0002).

In each study group patients whoreceived the influenza vaccinationwere compared to those who did notreceive it. The significant findings areillustrated in Table 2. Previous influ-enza vaccination and a greater numberof chronic illnesses were associated

with influenza vaccination in bothgroups. The presence of a chronic ill-ness was associated with vaccinationin the Drop-In group, and there was atrend toward significance in the Phonegroup. Increasing age and a greaternumber of visits to the FMC were sig-nificantly associated with flu vaccina-tion, but only in the Phone group. Noother associations were found.

Telephone survey resultsEighty-eight of the 120 patients in

the Phone group (73.3%) were con-tacted by telephone. Patients in thePhone group did not receive a call ifthey happened to visit the FMC beforeit was their turn to be telephoned (22of 120). Excluding these, the actualtelephone contact rate was 89.9%.Only 10 patients remained unnotifiedafter three attempts. A comparison ofthe patients who were contacted withthose who were not' showed no signif-icant differences in demographic orhealth factors.

In total, 132 phone calls (excludingbusy lines) were made by the threecallers. Most (77.6%) of the patientscalled were contacted on the first tele-

Table 1Influenza Immunization Rate for Those Age Eligible

% Immunized % ImmunizedPhone Drop in

Year Group Group P Value Overall1983 10.9(n=92) 8.2(n= 98) .7000 9.5(n=290)1984 7.3(n= 109) 4.5(n= 111) .5400 5.9(n= 220)1985 50.8(n= 120) 26.8(n= 123) .0002 38.7(n= 243)

CAN. FAM. PHYSICIAN Vol. 33: SEPTEMBER 1987 1999

phone call. The proportion of patientscontacted rose to almost 90% afterthree calls. In only three instanceswas a patient's phone busy.The project required relatively little

staff time. The physician made hiscalls over a six-week period, attempt-ing, on average, one call per day. Thenursing staff averaged two calls perday over three-week calling period.

There was no significant differencein the influenza vaccination rate ofthose patients called by the physician,the nurse, or the nursing assistant.There was, however, a trend towardshigher rates in those called by thenursing staff.

For the most part vaccination wasgiven by "appointment" on a visit tothe doctor. Twenty of the 94 patientsreceiving shots (21.3%) were vaccin-ated by the nursing "clinic". Thosecontacted by telephone made muchgreater use of the clinic (P= .0003).

DiscussionClearly, an immunization policy

substantially improves the provision ofinfluenza vaccination to the elderly.Immunization rates of 50.8% for thePhone group and 26.8% for the Drop-In group represent increases of 43.2%and 22.7% respectively over the pre-vious year's levels.A simple office immunization pol-

icy with no outreach measures, as re-presented by the Drop-In group, re-sulted in the immunization of overone-quarter of the study population.This policy approach, although in-creasing the immunization rate six-fold, was limited by its design. Onlythose visiting during the vaccination

period could receive the shot. In thisstudy, 63.4% of the Drop-In group vis-ited the FMC during the vaccinationperiod, and 42.3% of these were im-munized. Although this approach hasits limitations, it is easy to implementand entails no expense.

McDowell et al. obtained similarfigures for their "physician reminder"group. In this group the physician,rather than the nurse, as in our study,inquired about influenza vaccination.The participants obtained immuniza-tion rates of 22.9% overall, with a43.5% rate for those the doctor actu-ally saw. ' 6

Attempting to reach the entire at-risk population requires the use of out-reach measures. Sending postcards toremind patients about influenza vacci-nation has been extensively studied inthe past.1, 6-8, 14 Reminding patientsby telephone has been studied only re-cently by McDowell et al.,' who ob-tained a telephone-group immuniza-tion rate of 37% .16 The 50.8%immunization rate that was recorded inthis study has confirmed that the tele-phone is an effective alternative as aninstrument for raising the rates of in-fluenza immunization. The substantialincrease in vaccination coverageachieved by using the telephone in thisstudy is similar to that found in thesuccessful postcard-reminder studiescarried out in the United States andCanada. ", 8 The decision on which tac-tic to employ may thus depend on costfactors and office logistics.

Postcard studies are by their naturemore expensive to conduct. Materialand postage costs may be in the neigh-bourhood of 40 cents per letter.20 For

Table 2Demographic and Health Factor Associations by Study Group

* Factor Phone Group Drop in GroupPrevious InfluenzaVaccination P<.0001 P .001

Age (Older) P .01 N.S. (.763)Presence of aChronic Illness P=0.56 P .01

Number of ChronicIllness P-.05 P<.05

Mean Visits to P-.001 (1983) N.S. (1.67)FMC P<.01 (1984) N.S. (.649)

an elderly population of 250 patientsthis would mean an expense of Cdn.$100.00. The use of the telephone, onthe other hand, entails no added ex-pense, as the telephone is necessary tothe normal functioning of the medicaloffice.

Preparing and distributing postcardreminders require manpower hours.The telephone approach also requiresmanpower hours, but distributed dur-ing office time and shared by staffmembers, the actual burden is quitereasonable. Based on the telephonesurvey data, three people would haveto make only two telephone calls perday for six weeks to cover 250 elderlypatients. If a single nurse or secretarywere assigned the task, it would re-quire making six calls per day for sixweeks. Participants in this study foundthat the nursing staff and the physiciancould easily place their calls at lowwork periods during the day.The telephone approach may have

other advantages. The calling list doesnot depend on successful delivery ofthe mail or on a correct mailing ad-dress. Over 90% of the patients in thePhone group were contacted. Anyerrors in telephone numbers can becorrected immediately. Moreover thehuman element involved in telephonecontact would seem to have a deepereffect than an impersonal computer-printed message. The written messagerequires that the recipient have ade-quate vision and literacy; a phonecaller can judge the patient's under-standing by his/her response. The tele-phone also permits the patient to askquestions and get feedback.The telephone calls appeared to in-

fluence patients who were older,chronically ill, and visited the FMCmore frequently. The reason for thesepatients' higher response rate is uncer-tain. It may represent their greaterlikelihood of being contacted by tele-phone, or a stronger effect on personswith declining faculties, or their closerdependence on the FMC. In any eventthe telephone contact affected the typeof patient (the sick and aged) at great-est risk from influenza disease andthus in greatest need of influenza im-munization. A higher response rate byage selection process was not found inthe Drop-In group or in one postcardstudy that looked for it.' 5 The effect ofchronic illness on influenza-vaccineresponse rates has not been examinedin post card studies.

CAN. FAM. PHYSICIAN Vol. 33: SEPTEMBER 1987

N.S.=not significant (P >.05)* all other factors (sex, marital status, household composition, yearsattending FMC and prior adverse reaction) were N.S.

2000

The study determined that previousinfluenza immunization was a positivefactor affecting an individual's im-munization status in both groups. Thisfinding has been documented for thepostcard approach, also.15 It mayimply a general satisfaction with thevaccine. Moreover, it may indicate a"'carry-over" effect, suggesting that itis easier to ensure yearly re-immuniza-tion than initial immunization.The availability of the nurse to pro-

vide the shot at a time that was agree-able to the patient was useful. It servedthe Phone group predominantly, pro-viding patients with the option of re-ceiving the influenza shot without hav-ing to make a doctor's visit. It was alsouseful when vaccinations had to bepostponed because of a febrile illness.The telephone approach may be use-

ful in other areas of prevention such asPap-smear recall, annual health exam-inations, other immunizations such asthe pneumococcal vaccine, occultblood and signoid re-examination.A phone-call system might evenbe flagged by a daily computer prin-tout that notified staff of preventiveconcerns. The telephone represents afamiliar and readily available technol-ogy of which the preventive capabili-ties have been overlooked and un-derused.

AcknowledgementsWe are indebted to the following

people, all of whom played importantroles in making this research project asuccess: Ms. Susan Hoddinott for sta-tistical help and guidance; Dr. MartinBass for organizational help and inter-pretation; Dr. Sergio Lappano, Mrs.Harriet Mol, R.N. and Mrs. BarbaraSimani, R.N.A. for implementing thestudy protocols; and Mrs. Mary AnnLos, for typing (and retyping).

References1. Recommendation of the ImmunizationPractices Advisory Committee (ACIP).Prevention and control of influenza.MMWR 1984; 33:253-66.

2. Schoenbaum SC. Inflpenza vaccine-unacceptable or unaccepted (editorial). AmJ Public Health 1979; 69:219-2 1.

3. College of Family Physicians of Can-ada. Health maintenance guide. Willow-dale, Ont.: The College, 1983; 34-6.

4. Barker WH Mullooly JP. Influenzavaccination of eiderly persons: reduction inpneumonia and influenza hospitalizationsand deaths. JAMA 1980; 244:2547-9.

5. Ruben FL, Johnston F, Streiff EJ. Influ-enza in a partially immunized aged popula-tion: effectiveness of killed Hong Kongvaccine against infection with the Englandstrain. JAMA 1974; 230:863- 6.6. Briscoe JH. The protective effect of in-fluenza vaccine in a mixed influenza A andB epidemic in a boys' boarding school. J RColl Gen Pract 1977; 27:28-31.7. Impact of influenza on a nursing homepopulation-New York. MMWR 1983;32:32-4.8. Patriarca PA, Weber JA, Parker RA, etal. Efficacy of influenza vaccine in nursinghomes. Reduction in illness and complica-tions during an influenza A(H3W2) epi-demic. JAMA 1985; 253:1136-9.9. National Advisory Committee on Im-munization (NACI). Statement on influ-enza vaccination for the 1986- 1987 sea-son. Can Dis Weekly Report 1986; 12:77-84.10. Kavet J. Vaccine utilization: trends inthe implementation of public policy in theUSA. In: Selby P, ed. Injluenza; virus,vaccines, and strategy. New York: Aca-demic Press, 1976; 297-308.11. U.S. Department of Health, Education,and Welfare, Centers for Disease Control.United States immunization syrvey. At-lanta, Ga.: The Department, 1976. DHEWpublication no. (CDC) 78- 8221.12. Anderson C, Hodges M. Effectivenessof patient recall system on immunizationrates for influenza. J Fam Pract 1979;9:727- 30.13. Larson EB, Olsen ED, Cole WC, et al.The relationship of health beliefs and apostcard reminder to influenza vaccination.J Fam Pract 1979; 8:1207- 11.I14. Henk M, Froom J. Outreach by pri-mary-care physicians. JAMA 1975; 233:256-9.15. Frank JW, Henderson M, McMurrayL. Influenza vaccination in the elderly: 1.Determinants of acceptance. Can MedAssocJ 1985; 132:371-5.16. McDowell I, Newell C, Rosser W.Comparison of three methods of recallingpatients for influenza vaccination. CanMedAssocJ 1986; 135:991-7.

17. Department of Clinical Epidemiologyand Biostatistics, McMaster UniversityHealth Sciences Centre. How to read clini-cal journals. V-To distinguish usefulfrom useless or even harmful therapy. CanMed Assoc J 1981; 124:1156-62.

18. Barker WH, Mullooly JP. Pneumoniaand influenza deaths during epidemics: im-plications for prevention. Arch Intern Med1982; 142:85- 9.

19. Barker WH, Mallooly JP. lmpact ofepidemic type A influenza in a definedadult population. Am J Epidemiol 1980;112:798- 813.

-0. Frank JW, McMurray L, HendersonM. Influenza vaccination in the elderly: 2.the economics of study reminder letters.Can M/ed Assci 1/ 985; 132:51I6-8S, 5S21.

sTussionsex(resin complexes of hydrmcodone and phenyltoloxamine)

6 Days of Antitussive Therapy

Other Leading Antitussives*

'liii,Manufacturer's recommended dosage, eg.

ANCALAB PHARMACEUTICALSDOW PHARMACEUTICALS

sTussionex(rein complex d hydroodone and phenyoloxamine)

6 Days of Antitussive Therapyin a 60 ml prescription

MEMblER

IPAAB| PMAC

This product has a potential for being abused.

Full prescribing Information availableupon request.

E!F PEINWALTEPennwalt Inc., Pharmaceutical Division,

1851 Sandstone Manor,Pickering,Ontario L1W 3R9*Registered trademark of Penowali Corporation

CAN. FAM. PHYSICIAN Vol. 33: SEPTEMBER 1987 2001