SEARCH

SEARCH BY CITATION

Keywords:

  • practice guidelines;
  • intervention;
  • education (doctor);
  • case management;
  • feedback

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

Background Standard Treatment Guidelines were introduced to all prescribers at provincial hospitals in Lao PDR and treatment indicators were developed within the National Drug Policy programme to monitor compliance.

Objectives To evaluate the effects of an educational intervention to improve treatment practices of malaria, diarrhoea and pneumonia.

Methods Randomized controlled trial with prescribers at 24 departments at eight provincial hospitals, matched into four pairs. The three departments of internal medicine, paediatrics and out-patients in each pair were randomized into intervention or control group. The 6-month intervention was conducted by members of the Drug and Therapeutics Committees, and comprised monthly audit sessions in the form of outcome feedback using indicator scores on recorded treatment of malaria, diarrhoea and pneumonia. We measured treatment indicator scores 6 months after the end of the intervention compared with baseline.

Results The aggregated mean scores for all diseases, and for malaria and diarrhoea, improved significantly. For pneumonia, improvement was seen in both the intervention and control groups. Record keeping was improved for all three diseases. For malaria, there were improvements in recording patients’ history, and in frequency of microscope testing; for diarrhoea, regarding weight measurements, palpation of the fontanel for children under 2 years, and reduction of irrational use of anti-diarrhoeals and antibiotics; for pneumonia, in recording respiratory count, and reducing irrational use of anti-histamines and anti-cough medications.

Conclusions Audit–feedback systems to improve quality of care are feasible and effective also in hospital settings in low-income countries.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

Several studies have been conducted on methods to improve prescribing and to solve problems associated with inappropriate prescribing (Laing et al. 2001; Thomson O'Brien et al. 2001a). Guidelines are one way to support effective clinical practice, but only written information about guidelines appears to be of limited value in changing doctor's behaviour (Buntinx et al. 1993; Kanouse & Jacoby 1988; Freemantle et al. 2001). Face-to-face information, including feedback, is a more effective tool to bring about a change of doctors’ knowledge, attitudes and actual behaviour (Wahlström et al. 1997; Stalsby Lundborg et al. 1999; Thomson O'Brien et al. 2001a). Reviews of getting research findings into practice imply that lectures and distribution of educational materials alone have little or no effect on changing behaviour among health professionals (Davis et al. 1995; Bero et al. 1998). Continuing medical education contributes in improving doctors’ performance and, in some instances, health care outcomes (Davis et al. 1992). A combination of methods is necessary in order to get an impact on practice behaviour (Bero et al. 1998). Audit-based quality improvement involves practitioners in developing guidelines and setting targets, in the implementation of the guidelines and the evaluation of performance through feedback and peer group discussions (Thomas 1994; Thomson O'Brien et al. 2001a).

Lao People's Democratic Republic (PDR) is a low-income country with a population of 5.2 million people living in 17 provinces, each with a provincial hospital. There are also three central hospitals in the capital, and in total 124 district hospitals. The per capita income is 330 US$, and the total expenditure in the under-developed health sector is only 4 USD per capita (National Statistics Centre, State Planning Committee 2001). As in many low-income countries, malaria, diarrhoea and pneumonia greatly contribute to the disease burden and to a high mortality risk. This can be associated with inappropriate case management, polypharmacy and overall irrational drug use, which have also been reported from Lao PDR (Paphassarang et al. 1995). Many factors contribute to inadequate treatment (Soumerai & Avorn 1989; Hardon et al. 1991), including poor management, unethical marketing, and poor adherence to full treatment doses of certain drugs for certain diseases. Lack of well-functioning continuing medical education is a further obstacle to improvements (Davis et al. 1992; Thomson O'Brien et al. 2001c).

Prescribers in Lao hospitals have limited opportunity to be trained and to learn about new treatment options and new technologies. To address the problems of irrational use of drugs, several steps were taken within the implementation of the National Drug Policy (NDP), which was endorsed in Lao PDR in 1992 (Paphassarang et al. 1995). One step was the development of Standard Treatment Guidelines (STGs) through a participatory process involving around 100 doctors. The first STGs were developed for seven of the most common and burdening diseases (malaria, pneumonia, diarrhoea, parasitosis, dengue fever, tuberculosis and leprosy). The guidelines were printed and distributed to provincial and district hospitals in the beginning of 1999. The Ministry of Health had advised all provincial hospitals to establish Drug and Therapeutics Committees (DTCs), and to be instrumental in the NDP implementation, including use of practice guidelines.

To facilitate implementation of an audit-based system, performance indicators had been developed in the NDP implementation programme funded by Sida (Swedish international cooperation agency). Indicators were developed both for the rational use of drugs – adapted from WHO (Brudon-Jakobowicz et al. 1994) – and for the newly developed STGs (Johansson 1998). The STG indicators were tested by the Department of Curative Medicine at the Ministry of Health through the DTC at the central Mahosot Hospital in Vientiane. The indicators address both diagnostic and treatment aspects of the management (see Appendix for details). This study was developed to give evidence of the appropriateness of the next step in the audit cycle in the context of provincial hospitals in Lao PDR.

The study aimed to assess the effects of an educational intervention on management of three common diseases (malaria, pneumonia and diarrhoea), using performance indicators and an audit–feedback approach at selected departments at eight provincial hospitals.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

We performed a randomized, controlled trial with 24 selected departments at eight provincial hospitals as study units. The number of eligible hospitals was limited to these hospitals, as they were the only ones with functioning DTCs. Malaria, pneumonia and diarrhoea were selected as tracer conditions because of their high prevalence and large contribution to the disease burden. The study was conducted from June 1999 to August 2000 and included a pre-intervention period from June to September (STG introduction), an intervention period from September 1999 to February 2000, and a post-intervention period from March to August 2000.

The provincial hospitals of Luangprabang, Oudomxay, Khammouane, Savannakheth, Champassack, and Borikhamxay, and the Mittaphab and Setthathirat hospitals in the Vientiane municipality were selected on the grounds of having an established DTC or being in the process of establishing one. At the time of selection all of these were provincial hospitals that met the criteria for inclusion, including being comparable in matched pairs (see below). In each of the hospitals the out-patient department (OPD) (including emergency), the department of paediatrics, and the department of internal medicine (including infectious diseases) were selected as the three tracer diseases are commonly treated there. All prescribers (doctors and medical assistants) at the selected departments were included as participants.

A pair-wise balanced block design was used in the randomization process to control for differences in size (number of attendancies, number of beds, number of doctors and medical assistants), function (provincial or combined regional/provincial), and type of department. The eight hospitals were therefore grouped in four pairs in relation to size and function. Each pair of hospitals was thereafter randomly allocated to one of the four pairs in the balanced block. In each pair the hospitals were randomly assigned to one of the two blocks. The allocation of the three departments to either the intervention or control group had already been assigned in the block.

Through these procedures each hospital had either one intervention and two control departments, or one control and two intervention departments. In the whole study sample there were four intervention and four control departments for each type of disease. In total there were 12 intervention and 12 control departments serving as study units. Background characteristics of departments and prescribers are shown in Table 1. The number of beds and attendancies were of the same magnitude in both study groups. Eighty-six per cent of the 122 prescribers were women. There were no differences between the intervention and control groups regarding number, educational level and work experience of the prescribers, except that there were more medical assistants in the control group (39 vs. 26).

Table 1.  Background characteristics of departments (n = 24) and prescribers (n = 122) in eight provincial hospitals
 Intervention departments (n = 12)Control departments (n = 12)
Out-patient departments (OPD)44
Paediatric departments44
Internal medicine (infectious diseases) departments44
No. of beds for paediatric and internal medicine286329
No. of in-patients  
 Paediatric department12031190
 Internal medicine14281307
No. of patients attending OPD84159613
No. of medical doctors2730
No. of medical assistants2639
No. of woman prescribers/total number44/5360/69
No. of prescribers trained in Laos45/5361/69
No. of prescribers with working experience >5 years37/5343/69

Assessment of clinical performance

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

The clinical performance of doctors and medical assistants could not be observed directly. Instead, the existing STG indicators (see Appendix) were used to assess and compare selected aspects of the diagnostic and treatment components of case management. There were different indicators for out-patients and patients who were hospitalized. The source of information was the record keeping book. For each indicator, several components were assessed (see Appendix for details), and 30 patients should be selected to make the scoring reasonably representative. The data collection was performed by members of the DTC, who were familiar with the procedure as they had previously collected the same data on a routine basis. It was not possible to perform this collection in a blinded way, as there were no resources to recruit and train independent investigators.

The STG indicator scores for the selected diseases on recorded performance were collected from 3 months before the intervention, to 6 months after the intervention. The results of the assessment were presented as scores for each component, and a total indicator sum score for each disease. The components of the indicators were given different weight depending on their assessed importance in the clinical management. Negative scores were used for particularly inappropriate actions taken. For easy comparison the maximum score for each indicator was 10. Cases of multiple diagnosis were included if one of the target diseases was clearly indicated in the record book.

Intervention

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

There was a regular intervention in all departments in all hospitals comprising the introduction of the new STGs to all doctors and medical assistants during the pre-intervention phase (June–September 1999). Appointed DTC members at each hospital were encouraged to use the STG indicators every month to measure the doctors’ performance for all three diseases. This information was available for the respective departments, but, in the control departments, there was no active promotion.

At the intervention departments, feedback sessions with interactive discussions were to be carried out with all prescribers every month between September 1999 and February 2000, facilitated by the appointed senior DTC member. At each meeting only one of the diseases was addressed, but each disease should be presented and discussed at two separate occasions. The facilitators did analyses of the aggregated indicator scores for the previous 3 months in relation to the guidelines. They also agreed with other DTC members (including heads of departments) on the relevant messages for the group discussions with prescribers. Each session was planned to last 60–90 min, including reflection on the feedback, and discussions on how to improve performance.

The facilitators (five men, three women) were trained by the research group in appropriate STG management, and in how to lead a feedback session. The facilitators were trained by the Laotian members of the research group. The members of the research group were trained during workshops with support from the Swedish authors. The performance of the facilitators was assessed through observation by the Laotian members of the research group. These observations were communicated to the facilitators individually.

Outcome measurements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

The outcome of the intervention was measured through the STG indicators by comparing the intervention and control group regarding the mean sum scores of recorded performance of all prescribers in each department 3 months before the intervention, during the intervention, and 3 and 6 months after the intervention, for all diseases together (primary outcome) and for each of the selected diseases (secondary outcomes).

Research team members made regular visits to control and monitor the process to make sure that data were collected adequately. Feedback educational sessions were also attended in each provincial hospital, one at 2 months after the start of the intervention and one at the end of the intervention.

Data analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

To compare the changes in mean STG indicator scores between the intervention and control groups, we used the residuals from a simple linear regression of the post- and pre-differences on the pre-intervention values rather than the differences as such. This was done to eliminate the influence of variation in pre-intervention values on the comparisons and the effect of regression towards the mean. The residuals, hence, are the deviations from the ‘expected’ change given a certain start value.

The residual was the dependent variable in a general linear model with the binary variable intervention or control and hospital as independent variables. Of the totally 23 degrees of freedom (d.f.), one is for the intervention control comparison and seven for the hospitals leaving 15 d.f. for the error term. Each diagnosis was treated separately.

All data were entered in the Epi-Info programme and later transferred to the SAS program, in which all analyses were performed.

Ethics

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

All prescribers and DTC members at all hospitals accepted participation in the study. The study was approved by the Council of Medical Sciences, Ministry of Health in Laos.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

The intervention sessions were carried out in all hospitals for all diseases, varying between one and three for each disease in each hospital department. The total number of sessions for each disease was 16 for malaria, 13 for diarrhoea and 17 for pneumonia. Members of the research team observed some of the sessions. It was assessed that the DTC members functioned satisfactorily in their roles as facilitators, accomplishing their tasks by presenting the accurate feedback data and encouraging discussions among the prescribers.

Comparison of regular introduction of STG with intensive feedback discussions

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

Our primary outcome, the total mean score for all three diseases, increased from 6.16 in the pre-intervention period to 7.38 in the second post-intervention period for the 12 departments in the control group, and from 6.48 to 7.94 for the 12 departments in the intervention group. The secondary outcomes, treatment indicator scores of each disease in the control and intervention groups, respectively, are shown in Table 2, before, during and after the intervention. In the control group, the scores for malaria, diarrhoea and pneumonia increased from 6.75 to 7.88, from 5.36 to 6.83 and from 6.38 to 7.43, respectively. Corresponding figures for the intervention group were from 6.44 to 8.66 for malaria, from 5.91 to 7.73 for diarrhoea and from 7.08 to 7.43 for pneumonia.

Table 2.  Mean indicator scores of treatment for malaria, diarrhoea and pneumonia in the 24 departments in the control and intervention group
 Pre- interventionAfter 3-month interventionAfter 6-month intervention3 months post- intervention6 months post- intervention
  1. * Figures within brackets indicate the standard deviations.

Control
Malaria6.75 (2.08)*5.12 (1.94)7.43 (1.47)7.46 (0.98)7.88 (1.12)
Diarrhoea5.36 (1.64)5.40 (1.30)7.05 (1.33)6.46 (0.80)6.83 (0.90)
Pneumonia6.38 (1.40)7.00 (1.15)7.15 (0.79)7.33 (1.19)7.43 (1.31)
Intervention
Malaria6.44 (1.98)6.88 (1.97)7.53 (0.85)8.14 (0.60)8.66 (0.95)
Diarrhoea5.91 (1.53)6.33 (1.47)7.71 (1.69)7.10 (1.38)7.73 (0.60)
Pneumonia7.08 (0.73)7.19 (1.17)7.80 (0.60)8.18 (1.07)7.43 (1.49)

The difference in improvement for the mean score for all three diseases was statistically significant in favour of the intervention group (see Table 3 for details). The improvements in the intervention group were also statistically significantly higher for malaria and diarrhoea, when comparing with the control group (see Table 3 for details).

Table 3.  Mean differences between the second post-intervention and the pre-intervention STG indicator scores
DiagnosisDifference between mean residual changes95% CIP-value*
  1. * The increase in mean STG indicator value from the pre- to post-intervention period is statistically significantly larger in the intervention group than in the control group for all diseases and for malaria and diarrhoea.

All diseases (analysis of averages for all three diseases, total number of observations = 24)0.630.16–1.120.012
Diarrhoea0.860.28–1.420.006
Malaria0.650.21–1.130.006
Pneumonia0.35−0.52–1.210.407

In the intervention group the malaria score improved steadily during all periods, for pneumonia there was a steady improvement up to the last period, and for diarrhoea there was first an increase, then a slight reduction and then an increase again. In the control group the patterns of improvements and reductions in the scores were more mixed. The scoring of each component within one indicator was not recorded. However, in collaboration with the research group, these scores were analysed by the facilitators for presentation in the feedback sessions. The following presentation of more detailed findings is based on these analyses.

Typically case management of malaria patients in the pre-intervention period lacked some diagnostic components of the treatment indicator score such as measurement of weight, not recording patient's history, especially treatment taken before coming to the hospital, and not performing microscopy testing. All these components of case management improved in the intervention group, but scores for follow-up remained low.

The typical case management of a patient with diarrhoea in the pre-intervention period also lacked taking and recording the patient's history, measurement of weight, and palpating the fontanel in children under 2 years old. Furthermore, the prescribers used antibiotics and anti-diarrhoeal medication in nearly all cases, which was not recommended in the STG. Information on how to use oral rehydration solution (dosage, duration, etc.) also scored low. Six months after the intervention all these components were improved in the intervention group.

Case management of pneumonia patients in the pre-intervention period lacked recording patient's history and recording respiratory count, anti-histamine and anti-cough medication were used in nearly all cases, and patients were discharged before 4 days of admission without any follow-up. Improvements were seen after the intervention, but history taking and diagnosis still scored lower than acceptable. The control group also showed improvements, but to a lesser extent and less consistently. Health education as part of case management, was an area where improvement was seen in both groups.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

We found improved combined mean treatment scores for all three diseases in both the intervention and control group, statistically more improved in the intervention group. We also found statistically significantly higher treatment scores in the intervention group for malaria and diarrhoea. The improvements in the post-intervention period concerned several clinically significant aspects of case management as documented in the medical records: (1) a more adequate and complete use of the patient's record book-keeping system, (2) taking and recording of the patient's history, especially on treatment before coming to the hospital, (3) treatment practices using STG, especially dose and duration of medication, (4) blood testing on day 3 before discharge for malaria patients, (5) checking weight in diarrhoea patients, (6) counting respiratory rate in pneumonia patients, (7) giving health education, and (8) more concern about follow-up.

These findings indicate that the principles of quality assurance through an audit cycle approach can give clinically important improvement of case management of major diseases in hospital settings in a low-income country. The improvements in the control group indicate that also a more regular approach (establishment of DTCs, introduction of STGs, regular assessment of indicators) can be effective to some degree. These parts of the overall intervention can be said to be both partially managerial, and an example of ‘educational outreach’ (Thomson O'Brien et al. 2001b) or, in other terms, ‘academic detailing’ (Soumerai & Avorn 1990). However, the intervention in this study is a well-defined educational activity carried out mainly by the local facilitators, and not by facilitators from the authorities or the academy. Therefore, the results are most adequately interpreted as contributory evidence of an added value of introducing the audit cycle with intervention feedback in the clinical setting.

We found no reports of rigorous studies (randomized controlled trials or pre- and post-measurements with a control group) about educational interventions for these three diseases at hospitals in low-income countries. In a hospital-based, before-after designed study in Ghana and Jamaica (Wagaarachchi et al. 2001), a similar educational intervention based on an audit cycle was performed regarding management of obstetric emergencies. Significant improvements were recorded during a 12-month follow-up period. In one study in ambulatory care in South Africa the effects of a training workshop on good prescribing practices were evaluated in a randomized controlled trial regarding nurse practitioners’ prescribing for upper respiratory tract infections and diarrhoea (Meyer et al. 2001). Positive effects on prescribing indicators developed by WHO (Brudon-Jakobowicz et al. 1994) were found after a 3-month follow-up period. A study at health centers in Zambia also showed improved prescribing practices (Bexell et al. 1996).

A possible limitation of our study design is that it was not possible to fully separate the effects of the intervention to the intervention group as doctors sometimes worked at more than one department, and as they also met across departments both at the hospital and in private. This contamination bias would have a negative impact on the results, making it more difficult to show changes related to the active intervention. A similar potential effect was introduced by the fact that some prescribers in two hospitals worked in two departments (paediatric and OPD) due to shortage of staff. This contamination would also have a negative impact on the results. It should also be noted that the assessments of case management were based on recorded and not directly observed performance. This is, however, in line with most studies of a similar kind (Thorsen & Mäkelä 1999), and is related to the difficulties of designing a study where behaviour is observed at the same time without influencing behaviour inappropriately. As is most often the case in studies on educational interventions, the participants could not be blinded regarding their allocation to intervention or control group, as they obviously were informed about the study and knew whether they participated in the feedback sessions or not. This bias effect could have a positive effect on the results as the participants in the sessions could feel more recognized and could be considered to be more inclined to change behaviour, just because of this particular attention. However, as we study real practice, we do not consider this a major drawback in our study.

Our data supports previous findings that in order to effectively change practitioners’ behaviour, educational interventions should be continuously implemented because one particular intervention is unlikely to work with all practitioners (Grimshaw et al. 2001). In addition, regular clinical meetings would help prepare the ground for reinforcing improvement to become sustainable (Davis et al. 1999). The strengths of our study are that the follow-up period was after 6 months, that it was measured during the same time of the year as in the pre-intervention data, and that we have followed the trend through repeated measurements. Furthermore, the intervention only comprised elements that are feasible to include in regular hospital routines.

The assessment of the current practice of prescribers was conducted retrospectively using the treatment indicator sheet, and calculating the scores. For some departments, especially in OPD, both the intervention and control groups had low scores. There are probably several reasons, among which the following can be considered: (a) the record book was not standardized in all hospitals; (b) most doctors neither fully recorded all signs and symptoms, nor the dose of the prescribed drugs; and (c) some patients were mainly treated because of symptoms of other diseases and were not fully examined. It should also be emphasized that the indicators were developed within the context of the general development of medical practice in the hospitals, supported by the NDP programme. Performance indicators had previously been successfully developed and used in the private pharmacy sector, where a study on regulatory enforcement showed improved practices measured through the scores (Stenson et al. 2001).

Overall, the study shows that a systematically organized education programme with repeated feedback meetings improved the performance of prescribers at public hospitals, including the rational use of drugs. The record keeping system was also improved, and became a good source of all information needed for calculating the treatment indicator scores. The audit–feedback model developed for this study can be used systematically and become integrated into routine work, in order to improve the clinical performance of prescribers at provincial hospitals.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

This study was conducted with financial support from the Swedish International Development Co-operation Agency (Sida) and the National Drug Policy Program. We thank all other research team members (Dr Sommay Rattanavong, Dr Bounkhong Sisounthone, Dr Seun, Dr Somchan, Dr Phanasinh Sylavan, Dr Bouaphat) for their dedicated work, and Dr Ponemek Dalaloy, Minister of Health, Dr Boungnong Boupha, Director of National Institute of Public Health, Dr Sommone Phounsavath, Director of Curative Department, and all directors of the eight provincial hospitals (Dr Sichanh Himpaphanh, Dr Thongphat Vongprachit, Dr Somsanouk, Dr Phokham Prasithideth, Dr Khampho, Dr Bounchanh, Dr Eksavang Vongvichit, Dr Bouaphanh) for their valuable support. We are also grateful to all appointed DTC members in each hospital, and especially to Dr Rolf Johansson.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix
  • Bero LA, Grilli R, Grimshaw JM, Harvey E, Oxman AD & Thomson MA (1998) Closing the gap between research and practice: an overview of systematic reviews of interventions to promote the implementation of research findings. The Cochrane Effective Practice and Organization of Care Review Group. BMJ 317, 465468.
  • Bexell A, Lwando E, von Hofsten B, Tembo S, Eriksson B & Diwan VK (1996) Improving drug use through continuing education: a randomized controlled trial in Zambia. J Clin Epidemiol 49, 355357.
  • Brudon-Jakobowicz P, Rainhorn JD & Reich ME (1994) Indicators for Monitoring National Drug Policies. World Health Organization, Geneva.
  • Buntinx F, Winkens R, Grol R & Knottnerus JA (1993) Influencing diagnosis and preventive performances in ambulatory care by feedback and reminders. A review. Family Practice 11, 219228.
  • Davis DA, Thomson MA, Oxman AO & Haynes RD (1992) Evidence for the effectiveness of CME. A review of 50 randomized controlled trials. JAMA 268, 11111117.
  • Davis DA, Thomson MA, Oxman AD & Haynes RB (1995) Changing physician performance. A systematic review of the effect of continuing medical education strategies. JAMA 274, 700705.
  • Davis D, O'Brien MA, Freemantle N, Wolf FM, Mazmanian P & Taylor-Vaisey A (1999) Impact of formal continuing medical education: do conferences, workshop, rounds, and other traditional continuing education activities change physicians behaviour or health care outcomes? JAMA 282, 867874.
  • Freemantle N, Harvey EL, Wolf F, Grimshaw JM, Grilli R & Bero LA (2001) Printed educational materials: effects on professional practice and health care outcomes (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Update Software, Oxford.
  • Grimshaw JM, Shirran L, Thomas R et al. (2001) Changing provider behavior. An overview of systematic reviews of interventions. Medical Care 39 (Suppl. 2), II-2II-45.
  • Hardon A, Van der Geest S, Geerling H & Le Grand A (1991) The Provision and Use of Drugs in Developing Countries. Het Spinhuis, Amsterdam.
  • Johansson R (1998) Report on STG Indicators January 1998. IHCAR, Stockholm.
  • Kanouse DE & Jacoby (1988) When does information change practitioners’ behavior? International Journal of Technology Assessment in Health Care 4, 2733.
  • Laing R, Hogerzeil H & Ross-Degnan D (2001) Ten recommendations to improve use of medicines in developing countries. Health Policy Plan 16, 1320.
  • Meyer JC, Summers RS & Möller H (2001) Randomized, controlled trial of prescribing training in a South African province. Medical Education 35, 833840.
  • National Statistics Centre, State Planning Committee (2001) Basic Statistics about the Socio-Economic Development in the Lao PDR. National Statistical Centre, State Planning Committee, Vientiane.
  • Paphassarang C, Tomson G, Choprapawon C & Weerasuriya K (1995) The Lao National Drug Policy: lessons along the journey. Lancet 345, 433435.
  • Soumerai SB & Avorn J (1989) Improving drug prescribing in primary care: a critical analysis of the experimental literature. Milbank Quarterly 67, 268317.
  • Soumerai S & Avorn J (1990) Principles of educational outreach (‘academic detailing’) to improve clincial decision making. JAMA 263, 549556.
  • Stalsby Lundborg C, Wahlström R, Oke T, Tomson G & Diwan VK (1999) Influencing prescribing for urinary tract infection and asthma in primary care in Sweden – a randomised controlled trial of an interactive educational intervention. Journal of Clinical Epidemiology 52, 801812.
  • Stenson B, Syhakhang L, Stalsby Lundborg C, Eriksson B & Tomson G (2001) Private pharmacy practice and regulation. A randomized trial in Lao PDR. International Journal of Technology Assessment in Health Care 17, 579589.
  • Thomas S (1994) Standard setting in the Netherlands: impact of the human factor on guidelines development. British Journal of General Practice 44, 242243.
  • Thomson O'Brien MA, Oxman AD, Haynes RB, Davis DA, Freemantle N & Harvey EL (2001a) Educational outreach visits: effects on professional practice and health care outcomes (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Update Software, Oxford.
  • Thomson O'Brien MA, Oxman AD, Haynes RB, Davis DA, Freemantle N & Harvey EL (2001b) Audit and feedback: effects on professional practice and health care outcomes (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Update Software, Oxford.
  • Thomson O'Brien MA, Freemantle N, Oxman AD, Wolf F, Davis DA, Herrin J (2001c) Continuing education meetings and workshops: effects on professional practice and health care outcomes (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Update Software, Oxford.
  • ThorsenT & MäkeläM (eds) (1999) Changing Professional Practice. Theory and Practice of Clinical Guidelines Implementation. Danish Institute for Health Services Research and Development, Copenhagen.
  • Wagaarachchi PT, Graham WJ, Penney GC, McCaw-Binns A, Yeboah Antwi K & Hall MH (2001) Holding up a mirror: changing obstetric practice through criterion-based clinical audit in developing countries. International Journal of Gynecology and Obstetrics 74, 119130.
  • Wahlstrom R, Diwan VK, Tomson G, Oke T & Beermann B (1997) An educational experiment in primary care in Sweden: Improving the management of hypercholesterolaemic patients at risk. European Journal of General Practice 3, 96102.

Appendix

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Assessment of clinical performance
  6. Intervention
  7. Outcome measurements
  8. Data analysis
  9. Ethics
  10. Results
  11. Comparison of regular introduction of STG with intensive feedback discussions
  12. Discussion
  13. Acknowledgements
  14. References
  15. Appendix

Appendix 1

Table 4.  Indicators for measuring clinical performance (total maximum score = 10)
DiagnosisScore
Indicator for out-patient treatment of malaria 
Weight1
History 
 Treatment taken before coming1
  (asked and recorded in the file) 
History + symptoms (asked and recorded) 
 (a) headache 
 (b) fever (‘alternative’, of malaria type) 
 (c) chills 
 (d) from malaria area 
 (e) malaria season 
 (f) hepato-splenomegaly 
Examination 
 Three or more of (a)–(f)2
 Two of (a)–(f)1
Microscopy 
 Carried out and recorded2
 No microscopy−1
Treatment 
 Follows recommendation in STG2
 Correct dosage (weight)1
 Correct duration1
 Use of not recommended treatment−2
Indicator for treatment of hospitalized cases of malaria 
Weight0.5
History 
 Treatment taken before coming0.5
  (asked and recorded in the file) 
History + symptoms (asked and recorded) 
 (a) headache 
 (b) fever (‘alternative’, of malaria type) 
 (c) chills 
 (d) from malaria area 
 (e) malaria season 
 (f) hepato-splenomegaly 
 
Examination 
 Three or more of (a)–(f)2
 Two of (a)–(f)1
Microscopy 
 Carried out and recorded1
 No microscopy−1
Treatment 
 Follows recommendation in STG2
 Correct dosage (weight)1
 Correct duration1
 Use of not recommended treatment−2
Follow-up 
 The symptoms disappear (no fever)1
Microscopy 
 Check-up on day 30.5
 Check-up at discharge0.5
Indicator for out-patient treatment of pneumonia 
Weight1
History (asked and recorded in the file)0.5
 Cough0.5
 Fever0.5
 Treatment taken before coming0.5
Examination (examined and recorded in the file) 
 Respiratory rate increased1
 No tiraging + no cyanosis1
 Child not tired, not weak1
 Positive finding on stethoscopy1
Treatment 
 Treatment of other symptoms1
 Treatment follows recommendations in STG1
 Correct dose0.5
 Correct duration0.5
 Use of anti-cough medications or anti-histamine drugs−1
Indicator treatment of hospitalized cases of pneumonia 
Weight1
History (asked and recorded in the file)0.5
 Cough0.5
 Fever0.5
 Treatment taken before coming0.5
Examination (examined and recorded in the file) 
 Respiratory rate increased according to age group1
 Tiraging0.5
 Cyanosis0.5
 Child tired and weak1
 Positive finding on stethoscopy1
Treatment 
 Oxygen1
 Antibiotic following recommendations in STG1
 Correct dose0.5
 Correct duration0.5
 Discharge before 4 days−2
 Use of anti-cough medications or anti-histamine drugs−1
Indicator for out-patient treatment of diarrhoea 
Weight0.5
History (asked and recorded in the file) 
 No. of times of diarrhoea and how many days0.5
 No blood or mucus0.5
 No fever0.5
 Time for last urine output?0.5
 Treatment taken before coming0.5
Examination (examined and recorded in the file) 
 Tears OK + mouth not dry0.5
 Skin normal0.5
 Normal fontanel0.5
 Child not weak + not very tired0.5
Treatment 
 Oral rehydration solution1
 Correct dose and duration1
 Correct duration1
 Coconut fluid or breast milk1
 Use of antibiotics−1
 Use of anti-diarrhoeal drugs−2
 Education given + recorded1
Indicator for treatment of hospitalized cases of diarrhoea 
Weight0.5
History (asked and recorded in the file) 
 No. of times of diarrhoea and how many days0.5
 Diarrhoea with blood or mucus0.5
 Temperature (fever)0.5
 Time for last urine output?0.5
 Treatment taken before coming0.5
Examination (examined and recorded in the file) 
 No tears + dry mouth?0.5
 Skin abnormal?0.5
 Fontanel sunk?0.5
 Pulse high0.5
 Child weak + very tired0.5
Treatment 
 Oral rehydration solution1
 IV fluids1
 Correct dose/quantity (according to weight)1.5
 Other fluids (rice soup, coconut fluid and breast milk)1
 Use of antibiotics without bacterial identification−1
 Use of anti-diarrhoeal drugs−2