improving physician prescribing patterns to treat rhinopharyngitis. intervention strategies in two...

Pergamon 0277-9536(95)00398-3

Soc. Sci. &led. Vol. 42, No. 8, pp. 1185-1194, 1996 Copyright © 1996 Elsevier Science Ltd

Printed in Great Britain. All rights reserved 0277-9536/96 $15.00 + 0.00

IMPROVING PHYSICIAN PRESCRIBING PATTERNS TO TREAT RHINOPHARYNGITIS. INTERVENTION

STRATEGIES IN TWO HEALTH SYSTEMS OF MEXICO

R I C A R D O PI~REZ-CUEVAS,* H E C T O R GUISCAFRI~, O N O F R E MUlqOZ, H O R T E N S I A REYES, P A T R I C I A TOMI~, VITA LIBREROS

and G O N Z A L O GUTII~RREZ

Health Services Research lnterinstitutional Group: Secretaria de Salud, lnstituto Mexican, del Seguro Social, Ave. Cuauhtemol 330, CMN Siglo XXI, Blogue B Unidad de Congresos Col. Doctores 06726,

Mexico DF, Mexico

Abstract--To improve prescribing practices for rhinopharyngitis, an interactive educational intervention and a managerial intervention were carried out in 18 primary care facilities in metropolitan Mexico City. Four family medicine clinics of the Mexican Social Security Institute (IMSS) and 14 health centres of the Ministry of Health (SSA) were included. A quasi-experimental design was employed. One hundred and nineteen physicians (IMSS 68, SSA 51) participated. Sixty-five physicians (IMSS 32, SSA 33) were in the study group, while 54 were in the control group (IMSS 36, SSA 18). The study had four stages: (I) baseline, to evaluate the physicians' prescribing behaviour for rhinopharyngitis; (II) intervention, using an interactive educational workshop and a managerial peer review committee; (!II) post-intervention evaluation of short-term impact; and (IV) follow-up evaluation of long-term effect 18 months after the workshop. The control group did not receive any intervention but was evaluated at the same time as the study group. At baseline, most patients in both institutions received antibiotic prescriptions (IMSS 85.2%, SSA 68.8%). After the workshop, the percentage of patients receiving antibiotic prescriptions in the IMSS went from 85.2°/, to 48. 1%, while in the SSA it went from 68.8% to 49.1%. Appropriateness of treatment was analyzed using the physician as the unit of analysis. At baseline, 30*/0 of IMSS physicians in the study group treated their patients appropriately. After the intervention, this percentage increased to 57.7°/., and at the 18-month follow-up it was 54.2%. The SSA study group increased the appropriate use of antibiotics from 35.7% to ,16.2%, with this percentage falling to 40.9% after the 18-month follow-up period. In the control group there were no significant changes in prescribing patterns with respect to either the prescribing of antibiotics or the appropriateness of treatment. The intervention strategies were successful in both institutions. Forty per cent of physicians improved their prescribing practices after the workshop, with this change remaining in 27.5% of them throughout the follow-up period. On the other hand, 42.5% of the physicians did not change their prescribing practices after the intervention. The rest (I 7.5%) showed appropriate prescribing practices during all the stages of the study. We conclude that it is possible to improve the physicians" prescribing practices through interactive educational strategies and managerial interventions. This type of intervention can be an affordable way to provide continuing medical education to primary care physicians who do not have access to continuing educational activities, and to improve the quality of care they provide.

Key ,,ords---continuing medical education, antibiotics, appropriateness, educational research, quality of care

BACKGROUND

Inappropriate drug use to treat common diseases, such as acute diarrhoea or acute upper respiratory infection, has been consistently reported in the literature [I,2]. Three problems are commonly noticed: overprescription, where excessive amounts o f antibiotics are prescribed without justification [3,4]; omission, where products that should be prescribed are not prescribed (for example, oral rehydration solution to prevent dehydration in acute diarrhoea) [3,4]; and inadequate use [5, 6]

*Author for correspondence.

(for example, the use of inappropriate dosages). These problems highlight the use of poor clinical- therapeutic criteria for diagnosis and the lack of appropriate prescribing practices among physicians.

These deficiencies have been confronted in several ways, such as strengthening medical school curricula on the treatment of common diseases [7], setting up guidelines (algorithms, critical paths) [8] and support- ing activities designed to provide continuing medical education to practicing physicians [9, 10] who do not have access to educational activities.

In a previous study, our work group carried out an interactive educational intervention [11, 12] with

1185

1186 Ricardo Prrez-Cuevas et al.

family physicians. They were taught to improve their prescribing practices to treat acute diarrhoea and encouraged to criticize their own clinical performance [9, 10]. The educational strategy took place in two family medicine clinics of the Mexican Social Security Institute (lnstituto Mexicano del Seguro Social, IMSS), obtaining successful results. The family physicians improved the appropriateness of diarrhoea treatments by (a) increasing the use of oral rehydration solutions, (b) reducing the inadequate prescription of antibiotics, and (c) reducing the use of restrictive diets. Senior researchers at the IMSS instructed family physicians in this study [9, 10].

The success of the previous study encouraged us to carry out a similar educational strategy to improve physicians' prescribing practices in the treatment of rhinopharyngitis. This disease is the leading cause of visits in primary care settings [13, 14]. It is self-limited [15, 16], with easy and inexpensive treatment [17-20] and clearly defined criteria. Nevertheless, inappropriate treatment, such as the excessive prescription of antibiotics and symptomatic medicines, remains a serious problem t3, 41.

This study was carried out in primary care facilities of the two major health systems in the country. The Mexican Social Security Institute (IMSS) is a health system providing health care to workers and their families on a prepaid fee basis, covering almost 40% of the Mexican population. The other system, the Ministry of Health (SSA), is the governmental health system providing care to people with fewer resources and with no access to IMSS. It covers almost 30% of the Mexican population.

The study aimed to improve the rational prescribing of primary care physicians because of the many consequences of inappropriate antibiotic prescription, such as increase in antibiotic resistance, waste of medicines, lack of compliance, toxicity, rising costs and inadequate patient education.

In this paper we describe the methodology of the educational and managerial interventions and analyze the impact of the strategy on the prescribing practices of the physicians. We also discuss the relevance of continuing medical education for primary care physicians who provide services yet rarely participate in educational activities.

METHODS

This study was carried out from September 1990 to July 1992 in 18 primary care clinics located in the metropolitan area of Mexico City. It included four family medicine clinics of the IMSS and 14 health centres of the SSA.

Characteristics of the clinics

All of the clinics provide primary care services. Family medicine clinics of the IMSS have approxi-

mately 27 examining rooms each, and all these clinics have a clinical laboratory, X-ray laboratory and pharmacy. Health centres of the SSA are smaller clinics. There are one to twelve examining rooms in each, and only three of the fourteen clinics have a laboratory, X-ray lab and pharmacy.

Both institutions share an essential drug list regu- lated by the SSA. The observance of this list, however, is quite different between these two institutions. The IMSS strictly requires all physicians to use the essential drug list; it is a prepaid fee system facility in which patients receive medicines at no extra charge. On the other hand, the SSA is more flexible in the use of the essential drug list, but patients can buy drugs either in the clinic's pharmacy or elsewhere. Thus, the SSA physicians are less restricted to prescribing drugs included in the essential list.

Patients' demographic characteristics differ between the two health systems. The IMSS provides care to its enrollees and their families, and most of the services are provided to adults. The SSA provides services to people with scarce resources, mainly children.

Allocation of the physicians and the clinics

To carry out the study, a quasi-experimental design using a control group and a study group was employed [21,22]. One hundred and nineteen physicians (IMSS 68, SSA 51) participated in the study.

The study group consisted of 65 physicians, 32 working in two IMSS family medicine clinics and 33 working in nine SSA health centres. The control group consisted of 54 physicians, 36 working in two IMSS family medicine clinics and 18 working in five SSA health centres. Only physicians with permanent contracts working during the morning shift were included in the study.

Stages of the study

(I) Baseline stage. The baseline stage was carried out from September to December 1990. Physicians' prescribing practices were evaluated through a survey of patients with illness less than 21 days in duration, in whom the physicians had made a diagnosis of rhinopharyngitis. By rhinopharyngitis we mean the clinical picture characterized by runny nose and cough, with or without fever and with no concurrent infection. Between 8 and 15 patients were assessed for each physician.

To evaluate the physicians' prescribing practices, the information was gathered from several sources. Trained nurses interviewed patients, reviewed clinical records and reviewed prescriptions given by the physicians. These data were recorded in a question- naire which included information on duration of illness, clinical picture, length of visit, diagnosis and treatment given to the patient.

(II) Intervention strategies. The interactive educational intervention consisted of two phases: training of the instructors, and a workshop. The

Prescribing patterns to treat rhinopharyngitis 1187

managerial intervention consisted of the development of a peer review committee.

Training of the instructors. The instructors were four physicians from the staff of the study group clinics and four junior researchers from our work group (HR, PT, VL, RP). Those physicians were selected by the medical director of each clinic. They were chosen because of their interest and ability to participate in educational activities.

The training period lasted 40 hours. During this time the research project and current bibliography were reviewed and analyzed [16, 18, 20, 23, 24]. Find- ings of the baseline stage and an algorithm to treat rhinopharyngitis [8] were discussed.

Workshop. The workshop was carried out with the study group using interactive educational techniques [11, 12]. The study group was allocated to four sites; at each site there were one physician and one researcher as instructors. The workshop consisted of five daily 2-hour sessions over one week.

The following is a brief description of the activities in the five sessions:

(A) Presentation of the research project. The instructors assembled the clinic's physicians to inform them of and to discuss with them the purpose of the research. All of the clinic's physicians were invited to participate in the study. Most of them agreed to participate when they learned that it was an educational research project.

(B) Bibliographic review. Participating physicians received a current bibliography related to the appropriate treatment of rhinopharyngitis [16, 18,20,23,24]. At this time round-table discussions analyzing the appropriateness and applicability of the recommended treatments were organized.

(C) Baseline analysis. Data describing physicians' prescribing practices were presented to them. The results showed inappropriate prescribing practices. However, the group of physicians agreed with the findings and a discussion was carried out with them to analyze the results.

(D) Proposing the rhinopharyngitis therapeutic scheme. After the analysis of the physicians' prescribing practices, a therapeutic scheme for rhinopharyngitis, based on an algorithm pre- viously published by our research group [8], was proposed. The former algorithm was more complex, including criteria to treat other acute respiratory infections (tonsillitis, sinusitis, acute otitis media and bronchitis) besides rhinopharyngitis. After discussion with the physicians, the scheme was modified to make it simpler and more appropriate. Opinions and suggestions of the physicians to standardize the treatment criteria were considered. As a reinforcement strategy, the scheme was printed and posted in all the examining

rooms. All physicians were asked to use it voluntarily, but their decision to use it or not w a s respected.

(E) Discussion of clinical cases. To reinforce the use of the therapeutic scheme, eight scenarios illustrating common clinical situations were presented to the physicians. The group worked out the scenarios in round-table discussions using the proposed therapeutic scheme.

Peer review committee. The peer review committee was assembled to (a) strengthen and promote the use of the therapeutic scheme among the physicians, (b) evaluate how it was used in daily practice and (c) encourage the physicians to criticize their own clinical performance. It was formed by the instructors and two or three rotating physicians of the clinics. Every physician went to the meetings at least three times. The peer review committee met once a week and finished when all the medical staff had participated. In each session the members of the peer review committee reviewed clinical records of patients with rhinopharyngitis and evaluated the adherence of the physicians to the therapeutic scheme. With the peer review committee we encouraged a more partic- ipatory activity than with the workshop. The physicians reviewed clinical records, identified failures in the treatments, them and proposed improvement alternatives.

(IIl) Post-workshop evaluation. This evaluation was carried out over 3 months (April-June 1991), using the same data collection procedures as those used in the baseline stage. The group of nurses who had interviewed patients in the first stage served as interviewers for this stage as well. All the physicians in the baseline stage group were evaluated again,

(IV) Follow-up stage. This stage began between 12 and 18 months after the educational strategy had finished. The same criteria for data collection were followed. During the baseline and post-workshop stages, 119 physicians were evaluated, but in this stage only 79 of them (66.3% of the initial sample) remained. Twenty-five physicians (38.5%) from the study group and 15 physicians from the control group (27%) did not continue because of attrition, illness, changes of responsibilities in the clinics, etc. It is important to mention that none of the physicians refused to be evaluated at any stage.

Analysis

Some academic and professional characteristics of the physicians (Table 1) were analyzed using the chi-square and Mann-Whitney U tests [25]. The study and control groups of each institution were compared. Drugs prescribed during the baseline stage were analyzed (Table 2), comparing both institutions. The mean number of medicines per patient w a s

calculated using the number of drugs divided by the number of patients. The chi-square test was used to make the comparisons between institutions.

1188 Ricardo P~rez-Cuevas et al.

Table I. Study and control group physicians" characteristics

Characteristics

Mexican Institute Ministry of Health of Social Security

(SSA) (IMSS)

Control Study Control Study group group group group n = 1 8 n = 3 3 n = 3 6 n = 3 2

P value comparing

both institutions

Age (years) Median 36 35 43 43 < 0.0 I Interval (29-58) (28-68) (31-M) (31-58)

Postgraduate training None 88.8 84.8 33.3 40.6 <0.01 Family medicine 0.0 0.0 44.4 34.3 <0.01 Other spectahy I I. I 15. I 22.2 25.0 < 0.01

Years of practice Median 9 8 13 14 ns lnte[val (1-27) (1-28) (0-22) (0-21)

Private practice 33.3 30.3 19.4 18.7 <0.01 Work for other institution 5.5 27.2* 8.3 3.1 <0.01 Contact with pharmaceutical

company representative 94.4 93.9 25.0 31.2 < 0.0 I Non-medical remunerated

activities 5.5 6.0 13.8 9.3 ns

*P < 0.05 comparing between control and study group in Ministry of Health. ns = not significant.

The impact of the educational strategy after the workshop was measured separately in each institution, considering the physicians as the unit of analysis (Tables 3 and 4). All 119 physicians in the baseline and post-workshop stages were evaluated. The average proportion of appropriate treatments given during the baseline stage was compared to the average proportion of treatments given during the post-workshop stage. Antibiotics and symptomatic drugs were analyzed separately. A t-test [25] was used to evaluate the significance of the changes between the control and experimental groups.

The evaluation of the impact during the post-workshop and follow-up stages was done only for physicians we followed up on during all three stages of the study. They were evaluated by calculating the average proportion of antibiotics appropriately prescribed by each. Both institutions were analyzed in the same way (Figs 1 and 2).

The appropriateness of the treatment was evaluated according to the scheme of treatment proposed

Table 2. Drugs prescribed during baseline stage by the staff physicians

Social Ministry Security of Health (IMSS) (SSA)

Drugs n = 809 n = 544 prescribed (%) (%)

Antibiotics 85.2 68.8* Antivirals - - 4.6 Antipyretics 46. I 36.0* Cough drugs 49.0 40.2* Antihistamines 15.8 14.9 Other drugs m 4.6 23.8* No drugs 1.6 7.5* Mean no. of drugs per patient 2 2

n - Number of patients. 'Vitamins, bronchodilators, anti-inflammatory drugs. *P < 0.05.

during the workshop [I 8-20]. The basic criteria were as follows:

(A) Patients having less than 3 days of illness with runny nose, with or without fever, must not receive antibiotics as part of the treatment. Also, those patients with more than 3 days of illness without fever must be treated without antibiotics.

(B) Patients having more than 3 days of illness plus fever must receive antibiotics as part of the treatment. The antibiotics considered appropriate to prescribe were penicillin, ery- thromycin and cotrimoxazole. Prescribing any other antibiotic was considered inappropriate.

The statistical analysis of the post-workshop and follow-up stage was done in two different ways: the study and control groups were compared by the paired t-test [25], and prescribing practices among the three stages were compared by the Friedman test.

The physicians' behaviour was classified according to the appropriateness of the treatment they gave to the patients. A physician's behaviour was considered positive when he treated 50% or more of his patients appropriately and negative when he treated less than 50% of his patients appropriately. A positive change from one stage to another was a change from negative to positive behaviour, and a negative change was a change in the opposite direction. The significance of the changes was analyzed by Fisher's exact test [25].

The different combinations of prescribing behaviour, type of change and the persistence of the change throughout the different stages were classified as follows:

(I) Persistent positive behaviour: The physicians appropriately treated patients in the three stages of the study: baseline, post-workshop and follow-up.

(2) Stable positive change: Negative behaviour at

..= 50

u

~ 40 • -- O

~ 30

= ~ 20

~ ' ,0

~'~" 0

<

Prescribing patterns to treat rhinopharyngitis

[ ] C o n t r o l n = 12 ,,,- / /

- [ ] C o n t r o l n = 19 / / /

- - f / /

/ / /

f / /

f / /

* ' / /

/ / / / / /

/ / /

/ / /

B a s e l i n e s t a g e P o s t - w o r k s h o p

Study groups

' 77 ¢ / , ¢ / . ,

! - • . I / , r . , , , , ¢ / . ,

| ° . . , / I | " - " ¢ / ~

r , , , ,

F o l l o w - u p

( 1 8 m o n t h s )

, . .g o .~ 60 - * ,~ ~ [ ] Control n = 27 ~7~

._ Qa 50 - , " / /

.'o ~ [ ] Controln =21 / / /

• .= ~ 40 - / / / / / /

"~ 30 z / /

o ~ 20 , ' / / " ~ / / /

o '~ I 0 , , , / / • . ~ / / J

o ":" 0 z / . , ~ Baseline stage Post-workshop

< Study groups

Fig. 1. Effect of the intervention on the Ministry of Health (SSA) physicians. *P < 0.05.

1189

777 / f J / / J

f / /

/ / /

/ / /

/ / J

f / / / / / / / / f / J f / /

Follow-up (18 months)

Fig. 2. Effect of the intervention on the Mexican Institute of Social Security (IMSS) physicians. *P < 0.05.

baseline stage changing to positive behaviour in the two evaluation stages.

(3) Delayed positive change: Negative behaviour during baseline and post-workshop stages, changing to positive behaviour during the follow-up.

(4) Recurrent negative behaviour: Positive change dur ing pos t -workshop stage turning to negative behaviour dur ing the follow-up.

(5) Persistent negative behaviour: Dur ing all three stages the physician main ta ined negative behaviour. Less than 50% of the pat ients were treated appropriately.

Statistical analysis of the changes was performed using Cochran 's Q test [26].

At the end of the follow-up stage, the study group characterist ics were analyzed to see whether they had any relat ionship to the change in their prescribing behaviour (Table 6). The statistical analysis was performed using the chi-square and Mann- Whi t ney U tests.

RESULTS

Characteristics of the physicians

The characterist ics of the physicians from both inst i tut ions are shown in Table I. The physicians

working for the SSA were younger. Many of them were general practi t ioners with no pos tgraduate t ra ining and were working simultaneously in private practice or for ano ther public health insti tution. More than 90% received pharmaceut ica l representa- tives in the heal th centre, while less than 32% of IMSS physicians received such visits ( P < 0 . 0 1 ) . Most of the physicians working for the IMSS were family medicine specialists. Compared to the SSA medical staff, the IMSS staff worked more frequently in nonmedical paid activities (P <0.01) , such as business. After compar ing the characterist ics between groups in each insti tution, we found that many SSA study group physicians worked for o ther heal th insti- tut ions (P < 0.01).

Medicines prescribed during the bo.~eline stage

Analysis of the t reatment provided to each pat ient dur ing the baseline stage (Table 2) shows the high propor t ion of antibiotics prescribed within bo th inst i tut ions ( IMSS 85.2%, SSA 68.8%). SSA physicians prescribed antiviral drugs not ment ioned in the essential drug list but the IMSS physicians did not. In relation to prescription of symptomat ic drugs, IMSS physicians prescribed analgesics and cough drugs more frequently, while the SSA physicians

Table 3. Effect of the educational strategy on Ministry of Health (SSA) physicians' prescribing behaviour

Drugs prescribed

Average proportion of treatments prescribed

Baseline stage Postworkshop Percentage differences

Control Study Control Study . . . . . . . . . . n = 18 n = 33 n = 18 n = 33 Control Study

{% ) (%) (%) (%) (%) (%)

Antibiotics 97. I Only antibiotics 27.9 Antibiotics and symptomatics 69.2 Only symptomatics 21.3 Antipyretics 36. I Cough syrups 25.0 Antihistamines 12.8 No drug 9.3 Average no. of drugs per patient 2.0

80.7 86.5 62.7 - 10.6 -- 17.7" 16.2 19.8 13.6 --8.1 -2.6 64.5 66.7 49.1 -2.5 - 15.4" 23.9 26.2 31.9 4.9 + 8.0* 33.4 48.1 40.5 + 12.0" +7.1 48.7 38.2 38.8 + 13.2" - 9.9 16.1 12.9 16.1 +0.1 0.0 7.4 7.4 8.8 - 1.9 + 1.4 2.0 2.0 2.0

*P < 0.05. n = Number of physicians evaluated in ever)' stage.

1190 Ricardo Pc~rez-Cuevas et al.

Table 4. Effect of the educational strategy on Mexican Institute of Social Security (IMSS) physicians' prescribing behaviour

Average proportion of prescribed treatments

Baseline stage Postworkshop Percentage differences

Control S tudy Control Study Drugs n = 36 n = 32 n = 36 n = 32 Control Study prescribed (%) (%) (%) (%) (%) (%)

Antibiotics 95.7 88.0 96.9 56.3 1.2 - 31.7" Only antibiotics 7.7 7.9 10.1 8.2 2.4 0.3 Antibiotics and symptomatics 88.0 80.1 86.8 48. I - 1.2 - 32.0* Only symptomatic 10.5 17.5 I 1.6 48.5 I. I + 31.0" Antipyretics 50.5 42.6 50.8 48.5 0.3 5.9 Cough syrups 47.7 51.7 38.5 48.3 -9.2* -3.4 Antihistamines 17.7 19.7 20. l 26,8 2.4 + 7, I * No drug 1.6 4.9 1.5 3.1 -0.I - 1.8 Average no. of drugs per patient 2 2 2 2

*P < 0.05. n = Number of physicians evaluated in every stage.

prescribed vitamins, an t i - inf lammatory drugs and bronchodi la tors . There was a significant difference between the n u m b e r of IMSS and SSA pat ients who were not prescribed any drug ( IMSS 7.5%, SSA 1.6%) (P < 0.05). None the less, physicians f rom bo th inst i tut ions prescribed a mean of two drugs per patient.

Impact of the educational strategy on the use of drugs

The impact of the educat ional strategy was evaluated separately in each inst i tut ion. The SSA study group (Table 3) reduced the prescript ion of ant ibiot ics f rom 80.7% to 62.7% (P < 0 . 0 5 ) and increased the use of 'only ' symptomat ic drugs f rom 23.9% to 31.9% (P < 0.05). The SSA control group increased the use of analgesic and cough drugs. The IMSS study group (Table 4) reduced the use of ant ibiot ics f rom 80.1% to 48.1% ( P < 0 . 0 1 ) , increased the use of 'only ' symptomat ic drugs f rom 17.5% to 48.5% (P < 0.05) and increased the use of ant ih is tamines f rom 19.7% to 26.8% (P < 0.05). The control group did not change the use of ant ibiot ics bu t reduced the use of cough drugs.

Impact of the educational strategy on the appropriate use o f antibiotics

The SSA study group physicians (Fig. I) increased the appropr ia te use of ant ibiot ics f rom 35.7% to 46.2% after the workshop. This percentage had declined to 40.9% by the 18-month follow-up but was still significantly higher than tha t in the baseline stage (P < 0.05).

The IMSS study group physicians (Fig. 2)

increased the appropriateness of their treatments in a

substantial way. The proportion of physicians using

the scheme increased from 30% to 57.7% and after

the follow-up period was 54.2% (P < 0.05).

In the control group from both institutions, there

were no changes in antibiotic prescribing practices at

any of the stages (P > 0.05).

Change in the prescribing practices of physicians after the intervention

At the end of the study, the prescribing practices of the physicians f rom the study group who had com- pleted the three stages of the study were analyzed (Table 5). Only 17.5% of them had persistent positive

Table 5. Changes in antibiotics prescribing behaviour after the educational strategy

Control group Study group n =39 n =40

Type of prescribing behaviour n (%) n (%)

Positive persistent' behaviour 1 2.65 7 17.5 Stable b positive change c 0 0.0 I0 25.0 Delayed d positive change 5 12.8 I 2.5 Positive change and

negative recurrence ~ I 2.6 5 12.5 Negative persistent behaviour 32 82.0 17 42.5 Cochran's Q Test Q = 7 P > 0.05 Q = 22 P < 0,0001

Positive = > 50% of patients were treated according to scheme. Negative = < 50% of patients were treated according to scheme. n = Number of physicians. Evaluation stages: (i) baseline, (ii) post-workshop, (iii) follow up. "Persistent: without change in the three stages. bStable: without changes between stages (ii) and (iii). ~Change P < 0.05 by Fisher test among the stages (i) and (ii) or (ii) and Off). d Delayed: positive change between stages (ii) and (iii). eRecurrence: negative change between stages (ii) and ( i i i ) .


behaviour, and 40% had positive change after the workshop (P < 0.001). However, 12.5% of the 40% had negative recurrent behaviour, returning to prescribing antibiotics inappropriately. On the other hand, 42.5% of the physicians maintained persistent negative behaviour even after they had participated in the educational strategy and the peer review committee.

In the control group, only one physician maintained persistent positive behaviour. We observed that 12.8% of the physicians had a delayed positive change not due to the educational strategy. The rest of the staff (82%) had persistent negative behaviour (P > 0.05) during the three stages of the study.

Analysis of changes in prescribing practices according to physicians' characteristics

At the end of the follow-up stage, the characteristics of the physicians who maintained stable positive change and those with persistent negative behaviour were compared (Table 6). The analyzed physicians' characteristics were similar to those shown in Table 1. This analysis was performed to find associations between the physicians' academic and professional characteristics and their prescribing behaviour and to obtain a profile of a physician who might respond positively to continuing medical education activities. The analysis showed no associations between physician characteristics and their prescribing practices. However, when we analyzed by the institutions where the physicians worked, we found that 11 SSA physicians but only 6 IMSS physicians maintained negative behaviour. Among the 10 physicians who exhibited a persistent positive change, 9 of them were from the

IMSS staff and only one from the SSA staff (P < 0.01).

D I S C U S S I O N

This study was carried out in two health institutions that are quite different in their organization, procedures of providing services and type of population served. Nevertheless, these institutions share a severe deficiency in providing continuing medical education activities or research activities related to health services at the primary care level. We were encouraged to set up an interactive educational technique and a managerial intervention for primary care physicians for several reasons. Previous works have highlighted inappropriate prescribing practices to treat common diseases such as acute diarrhoea [3,5,6] or acute upper respiratory infections [3, 4]. This inappropriateness is reflected not only in high rates of antibiotic prescribing but also in the incorrect way the antibiotics are used; mistakes in dosage, selection and duration of use are frequently reported [27}.

Several factors have been considered to influence physicians to prescribe medicines inappropriately: (a) Shortcomings in medical education during both the undergraduate and the postgraduate training periods. The predominant educational trend supports prescribing drugs to all patients [6, 28]. (b) Lack of trustworthy clinical and pharmacological judgement. This problem has been stressed given the inappropriate prescribing practices that have been observed even for the treatment of mild and self-limited diseases. (c) Unreliable sources of pharmacological

Table 6. Relation between behaviour changes and some selected characteristics of physicians study group

Comparison between the physicians who changed and

those who did not

Stable Persistent positive negative

Selected change behaviour characteristics n = 10 n = 17 P value

Sex Male 6 7 ns Female 4 10 ns

Age Median 43 36 Quartiles (25-75%) (40-54) (31-44)

Postgraduate training None 6 I I Family medicine 0 4 ns Other specialty 4 2 ns

Private practice 4 4 ns Contact with pharmaceutical

company representative 8 14 ns Teaching activities I 0 ns Administrative positions 3 3 ns Institution

Ministry of Health (SSA) I 11 <0.01 Social Security (IMSS) 9 6

ns - not signif icant.

1192 Ricardo Prrez-Cuevas et al.

information about the properties and indications of drugs. Most physicians get their up-to-date information mainly from the intrusive advertising of pharmaceutical companies. They rely more on pamphlets and brochures from the drug companies [29-31] than on critical reading of responsible sources of information such as medical journals. (d) Pressure exerted by patients to receive drugs after the visits [31, 32]. (e) Desire o f the physicians to "do something' [31], and the ease of using medicines to fulfil this desire. (f) Lack of research and continuing medical education among primary care physicians to improve the quality of medical care they provide [32, 33].

The most important of all these factors is the lack of trustworthy clinical and pharmacological judgement. This deficiency is related at least partly to the current educational trend in schools of medicine where the physicians receive hospital-based training. Physicians' knowledge about the diagnosis and treatment of most common diseases is therefore vulner- able and biased [28]. Paradoxically, after the school training period, most of them will work in primary care facilities and never in hospital settings.

The main purpose of these primary care facilities is to treat the most common diseases, besides providing a high volume of services. Consequently, there is no infrastructure, resources, trained personnel or even time to carry out continuing medical education programs or research. As a result, these types of activities are seldom done. For most primary care physicians, continuing education and research are passive activities in which the physician listens to the presentations and barely participates. Unfortunately, moreover, most of the lectures are unrelated to the problems these physicians face in the examining rooms [34-36].

In this project, in contrast, the intervention strategies included several interactive educational and managerial features: (a) Feedback on the physicians' prescribing practices. During the workshop, both physicians and instructors analyzed the main results of the baseline stage. Physicians responded positively when confronted with their clinical performance deficiencies. Many of them were motivated to change and took part actively during the workshop and peer review committee. (b) Review of recent literature [36]. As part of the educational strategy, round-table discussions were held to analyze medical articles focused on the current treatment of rhinopharyngitis. (c) Group-oriented education [I 0, 37, 38]. Through the workshop and peer review committees, the physicians were encouraged to analyze their clinical performance and to judge their prescribing practices. These activities were designed to motivate them to improve their prescribing. (d) Printed reminders [39]. The therapeutic scheme was printed and posted in all the examining rooms. It was maintained until the study was concluded.

The positive impact of these strategies, in both health care institutions, after the 18 months' follow-

up can be supported by several points. Most of the physicians agreed to participate in the workshop and peer review committee. For many of them, to participate actively in round-table discussions and peer review committee activities represents a good opportunity to analyze their own clinical performance and to interact with other physicians. The organization of the daily activities in the clinics fosters isolation of the physicians rather than interaction among them. If they are encouraged to participate and to discuss clinical problems among themselves, the peer review committee may be a good approach to appraise the quality of care they provide. The physicians' own review of their practice through feedback information, as a problem-solving approach, has proved to be successful when it is set up along with decision-makers [40]. Particularly in primary care settings, the peer review committee is an unusual activity that could be implemented as part of the routine activities of the physicians. Providing feedback information about practices and performance can promote self-criticism and interaction among members of the staff and can improve their attitude toward the changes the organization wants to set up.

At the end of the study, most of the physicians who participated in the intervention were using appropriate treatments. There was a positive change after the workshop among 40% of the physicians. The change persisted in 27.5% of them after the follow-up period. However, 12.5% returned to inappropriate prescribing practices. The improvement in prescribing practices among approximately half the physicians supports the effectiveness of this type of intervention. The results are compatible with the results of other studies where similar educational strategies were used [9, 10]. Another important point is that 42.5% of the physicians maintained their negative behaviour throughout the study. We did not find differences in academic and professional characteristics between physicians who followed the scheme and those who did not follow it (Table 6). The only difference we found was the institution they worked for. More SSA than IMSS physicians maintained negative behaviour, but we do not have enough information to explain this finding. The percentage of physicians who did not change their prescribing behaviour was similar to the results obtained when the intervention was applied using acute diarrhoea as a study model. In that study 40% of the physicians exhibited negative persistent behaviour [10].

It is possible that there are other factors not studied that may influence behavioural changes. Primary care physicians rarely take part in activities other than clinical activities. They rarely participate regularly in continuing education activities. The way this medical system is organized to provide a high volume of curative services gives little or no opportunity to set up continuing educational activities. Therefore, it is difficult to improve the clinical performance of the


medical staff if the system does not support activities directed to this purpose. The routine of their clinical work contributes to the physicians' lack of commit- ment and weak response to educational activities. The physicians who participate regularly in continuing medical education activities are those who work in a hospital environment, have had postgraduate training and have teaching, research or administrative responsibilities as part of their clinical work [41]. We do consider that primary care physicians can participate actively in these activities besides their clinical work if they have enough support from the institutions they work for.

The purpose of the intervention was to influence prescribing habits positively. As mentioned above, multiple factors other than the academic background of the physicians influence prescribing practices. Some factors that have been suggested to be exam- ined when the goal is to influence prescribing practices are attitudes toward prescribing, normative beliefs, motivation, competencies and patient demand [421.

The prescribing behaviour of physicians is harder to modify because they are naturally inclined to do things in their own way. This autonomy is inherent in the professional context in which they work. Combining educational and managerial activities can facilitate mutual learning and promote cooperation rather than imposing coordination [43].

These types of intervention strategies have the advantages that they can be carried out in a few days and can be used in groups of 15-20 physicians [9, 10]. However, it is important for us to emphasize that other educational strategies should also be sought to motivate physicians toward more rational prescribing practices [44-46].

Another educational strategy we are testing is the clinical teaching centre. This centre was set up to promote appropriate case management of acute respiratory infections and acute diarrhoea [47-49] using a more comprehensive approach. It is located in a primary care clinic, where instructors, while providing clinical services, carry out face-to-face educational activities with physicians under training. The main feature of this clinical teaching centre is the focus on practising physicians. Physician trainees participate with medical staff and patients as they learn how to appropriately treat acute diarrhoea and acute respirator infections. This type of facility was intended for physicians who do not have access to continuing medical education. However, it is far more expensive to carry out this type of educational activity, because only up to six physicians in each round can be trained. Thus the time needed to instruct several physicians is much greater than the time needed to set up the workshop and the peer review committee.

A more affordable approach might be to combine both strategies, using the clinical teaching facilities to train physicians who then would be instructors in

interactive workshops [49]. This could be an alterna- tive way to improve the quality of the educational activities while educating more physicians in a cost- effective way.

We conclude that it is possible to change the inappropriate prescribing practices of physicians through interactive educational strategies. The like- lihood of this change may be related to the organiz- ational characteristics of the health institutions where they work more than their individual or professional characteristics. This type of strategy, combining interactive educational techniques and managerial interventions, can be considered an affordable means to provide continuing medical education to all working physicians to improve the quality of their prescribing.

Acknowledgements--Financial support for this research was provided by the Applied Diarrheal Disease Research Project at Harvard University by means of a Cooperative Agreement with the U.S. Agency for International Development.

REFERENCES

1. Soumerai S. B., McLaughlin T. and Avorn J. Improving drug prescribing in primary care. A critical analysis of the experimental literature. Milbank Q. 67, 268, 1989.

2. Soumerai S. B. and Ross-Degnan D. Drug prescribing in pediatrics: Challenges for quality improvement. Pediatrics 86, 782, 1990.

3. P~rez-Cuevas R., Mufioz O., Guiscafr6 H., Reyes H., Tom6 P., Libreros V. and Guti~rrez G. Patrones de prescripci6n terap~utica en infeccirn respiratoria aguda y diarrea aguda en dos instituciones de salud (IMSS. SSA); IV: Caractefisticas de la prescripci6n mrdica. Gac. Med, Mex. 128, 531, 1992.

4. Gutirrrez G., Martinez M. C. and Guiscafr6 H. Encuesta sobre el uso de antimicrobianos en infecciones respiratorias agudas en la poblaci6n rural me×icana. Bol. Med. Hosp. lnfan. Mex. 43, 761. 1986.

5. Mufioz O., Guiscafr6 H., Bronfman M. and Gutirrrez G. Estrategias para mejorar los patrones terap~uticos utilizados en diarrea aguda en unidades de atencirn mrdica primaria. IV. Caracteristicas del tratamiento prescrito pot los m&licos familiares y los pacientes. Arch. Invest. Med. Mex. 19, 371, 1988.

6. Gutirrrez G. El uso injustificado de medicamentos, un problema grave en aumento. Arch. Invest. Med. Mex. 19, 329, 1988.

7. Carruthers G., Goldberg T. and Segall H. Drug Utiliz- ation: A Comprehensive Literature Review. University of Toronto, Ontario, Canada, 1987.

8. Guiscafr6 H., Mufioz O. and Gutirrrez G. Normas para el tratamiento de las infeeciones respiratorias agudas. Bol. Med. Hosp. lnfan. Mex. 44, 58, 1987.

9. Guiscafr6 H., Mufioz O., Padilla G. and Reyes R. M. Estrategias para mejorar los patrones terap~uticos utilizados en diarrca aguda en unidades de atencirn mrdica primaria. VI. Evaluaci6n de una estrategia didgida a los m&licos familiares para incrementar el uso de la hidratacirn oral y disminuir el de antimicrobianos y dietas restrictivas. Arch. Invest. Med. Mex. 19, 395, 1988.

10. Gutirrrez G., Guiscafr~ H., Bronfman M., Walsh J., Martinez H. and Mufioz O. Changing physician prescribing patterns: Evaluation of an educational strategy for acute diarrhea in Mexico City. Med. Care 32, 436, 1994.

1194 Ricardo P6rez-Cuevas et al.

I I. Viniegra L. Anfilisis y perspectivas de la formaci6n de investigadores en el /Lrea de la salud. Ciencia 36, 231, 1985.

12. Viniegra L. La investigaci6n como herramienta de aprendizaje. Rev. Invest. Clin.(Mex.) 40, 191, 1988.

13. Clyde W. A. and Denny F. Acute respiratory tract infections: An overview. In Workshop on Acute Respir- atory Diseases Among Children of the World. University of North Carolina Press, Chapel Hill, 1983.

14. Direcci6n General de Estadistica y Proyectos Estrat6gi- cos. Anuario Estadistico. Secretaria de Salud, M6xico, 1990.

15. Acute Respiratory Infections in Children. Pan American Health Organization, ref RD 21/1, Washington DC, 1983.

16. Organizaci6n Panamericana de la Salud. lnfeeciones Respiratorias en los Ni~os: Manual para Mddicos. OPS, 1989.

17. Stansfield S. and Shepard D, Health Sector Priorities Review. Acute Respiratory Infections. Population, Health and Nutrition Division, Population and Human Resources Department, The World Bank. Oxford University Press, 1991.

18. World Health Organisation Case Management o f Acute Respiratory Infection in Children for Developing Countries. Manual for Doctors and Other Senior Health Workers. Acute Respiratory Infections Control Pro- gramme, World Health Organization, Geneva, 1989.

19. Manual de Normas para el Tratamiento de las lnfec- clones Respiratorias Agudas en los Ni~os. Secretaria de Salud, M6xico, 1991.

20. Household Management of Diarrhea and Acute Respir- atory Infections. Report of a scientific meeting at the Johns Hopkins School of Hygiene and Public Health. Control Programmes, World Health Organization, Geneva, 1991.

21. Castro Luis Dise$o Experimental sin Estadistica. Ed. Trillas, Mexico, 1986.

22. Varkevisser C., Pathmanathan I. and Brownlee A. Designing and Conducting Health Systems Research Projects, pp. 195-220. International Development Research Center, Ottawa, Canada, 1991.

23. Noticias sobre IRA. AHRTAG (WHO), 7, 1988. 24. Noticias sobre IRA. AHRTAG (WHO), 4, 1987. 25. Kirkwood B. Essentials of Medical Statistics. Blackwell,

Oxford, UK, 1988. 26. Siegel S. Estadistica No Paramdtrica. Ed. Trillas,

Mexico, 1988. 27. Duran L., Frenk J. and Becerra J. La calidad de la

conducta prescriptiva en atenci6n primaria. Salud Publica 32, 181, 1990.

28. Gutidrrez G., Guiscafrd H. and Mufioz O. Estrategias para mejorar los patrones terap6uficos utilizados en diarrea aguda en unidades de atenci6n medica primaria. X. Conclusiones y perspectivas de estudio. Arch. Invest. Med. Mex. 19, 437, 1988.

29. Moncada B. and Acevedo-Oliva B. El medico y la industria farmac~utica. Gac. Med. Mex. 4, 343, 1990.

30. Soumerai S. B. Factors influencing prescribing. Aust. J. Hosp. Pharm. 3, 9, 1988.

31. Schwartz R., Soumerai S. B. and Avorn J. Physician motivations for nonscientific drug prescribing. Soc. Sci. Med. 28, 577, 1989.

32. Greenberg R. A., Wagner E. H. and Wolf S. H. Physician opinions on the use of antibiotics in respiratory infections. J. Am. reed. Assoc. 7, 650, 1987.

33. Guti6rrez G., Bronfman M, Martinez M. C., Padilla G. and Mufioz O. Estrategias para mejorar los patrones teral~uticos utilizados en diarrea aguda en unidades de atenci6n medica primaria. 1. Metodolo- gia y caracteristicas de 1as unidades m6dicas y de la poblaci6n estudiada. Arch. Invest. Med. Mex. 19, 335, 1988.

34. Libreros B., Guiscafr6 H., Tome P., Reyes H., P6rez- Cuevas R. and Guti6rrez G. Patrones de prescripci6n terap6utica en infecci6n respiratoria aguda y diarrea aguda en dos instituciones de salud (IMSS, SSA). I. Metodologia del estudio y caracteristicas de las unidades m6dicas, de los m6dicos y de la poblaci6n estudiada. Gac. Med. Mex. 128, 505, 1992.

35. Viniegra L. Los intereses acad6micos en la educaci6n medica. Rev. Invest. Clin. (Mex.) 39, 281, 1987.

36. Gehlbach S., Wilkinson W., Hammond W., Clapp N., Finn A., Taylor W. and Rodell M. Improving drug prescribing in a primary care practice. Med. Care 22, 193, 1984.

37. Schaffner W., Wayne A., Federspiel C. and Miller W. Improving antibiotic prescribing in office practice. J. Am. reed. Assoc. 250, 1728, 1983.

38. Avorn J., Chen M. and Hartley R. Scientific versus commercial sources of influence on the prescribing behavior of physicians. Am. J. Med. 73, 4, 1982.

39. Tierney W., Hui S. and McDonald C. Delayed feedback of physician performance versus immediate reminders to perform preventive care. Med. Care 24, 659, 1986.

40. Mugford M., Banfield P. and O'Hanlon M. Effects of feedback of information on clinical practice: A review. Br. reed. J. 303, 398, 1991.

41. Shahabudin S. H. and Edariah A. B. Profile of doctors who participate in continuing medical education in Malaysia. Med. Educ. 25, 430, 1991.

42. Divine G. W., Brown J. T. and Frazier L. M. The unit of analysis error in studies about physicians' patient care behavior. J. Gen. Intern. Med. 7, 623, 1992.

43. Reinke W. A. Health Planning for Effective Manage- ment. Oxford University Press, New York, 1988.

44. Soumerai S. B. and Avorn J. Principles of educational outreach (academic detailing) to improve clinical decision making. J. Am. reed. Assoc. 263, 549, 1990.

45. Soumerai S. B., McLaughlin T. J. and Avorn J. Quality assurance for drug prescribing. Qual. Assur. Hlth Care 2, 37, 1990.

46. Avorn J. and Soumerai S. B. Improving drug therapy decisions through educational outreach. N. Engl. J. Med. 308, 1457, 1983.

47. Unidad de capacitaci6n para el tratamiento de la diarrea. In Dialogo sobre la Diarrea. AHRTAG (WHO) Vol. 37, 1991.

48. Patwari A. K. Cost-effective strategy for promotion of appropriate case management of diarrheal diseases. Establishment of DTUs (diarrhea treatment and training units). Indian J. Pediatr. 58, 783, 1991.

49. Bojalil R., Guiscafre H. and Gutierrez G. Centros docenteasistenciales para capacitacion a medicos en el manejo de las infeccioncs respiratorias agudas y la diarrea aguda. Grupo interinstitucional de Investiga- cion en Sistemas de Salud. Instituto Mexicano del Scguro Social, Sccretaria de Salud, Mexico (unpub- lished data), 1993.

improving physician prescribing patterns to treat rhinopharyngitis. intervention strategies in two...

Documents