Standard management of acute respiratory infections in a children's hospital in Pakistan: impact on antibiotic use and case fatality.

Standard management of acute respiratory infections in a children's hospital in Pakistan: impact on antibiotic use and case fatality* S.A. Qazi,1 G.N. Rehman,' & M.A. Khan2 Acute respiratory infections (ARI) are a leading cause of childhood morbidity and mortality in Pakistan. The National ARI Control Programme was launched in 1989 in order to reduce the mortality attributed to pneumonia, and rationalize the use of drugs in the management of patients with ARI. WHO's standard ARI case management guidelines were adopted to achieve these objectives. The medical staff at the Children's Hospital, Islamabad, were trained in such management in early 1990; further training sessions were conducted when new staff arrived. Data on outpatients were obtained from special ARI abstract registers, which have been maintained in the outpatient department since January 1990. Details on inpatients who were admitted with ARI were obtained from hospital registers. During the period 1989-92, the use of antibiotics in the outpatient depart- ment decreased from 54.6% to 22.9% (P < 0.0001). The case fatality rate (CFR) in children admitted with ARI fell from 9.9% to 4.9% (P < 0.0001), while the overall case fatality rate fell from 8.7% to 6.2%. Our results from a tertiary health care facility show that standard ARI case management reduced both antibiotic use and expenditure on drugs. Although the ARI case management criteria, which are more sensitive than the conventional diagnostic criteria of auscultation and radiography, led to more admissions, we believe that this strategy contributed to a significant reduction in the ARI case fatality rate. Introduction Among under-5-year-olds in Pakistan, acute respira- tory infections (ARI) are responsible for more than one-quarter of deaths in the community (1) and one- third of deaths occurring in hospitals (2). ARI is also the leading cause of morbidity in childhood.a Factors contributing to complications and mortality due to ARI include delays in taking the child to a health care provider, misdiagnosis or delay in diagnosis by the health professional (2), and inappropriate use of antibiotics for treatment of ARI (3-6). In 1989 the government of Pakistan launched the National ARI Control Programme, with the primary objective of reducing mortality from pneu- monia in under-5-year-olds and of rationalizing the use of drugs in ARI case management.a As part of * From the Federal ARI Cell, National ARI Control Programme, Children's Hospital, Pakistan Institute of Medical Sciences, G-8/3, Islamabad, Pakistan. Correspondence should be sent to Dr S.A. Qazi at this address. 1 Associate Physician, Children's Hospital, Islamabad, Pakistan. 2 Head of Paediatric Medicine Department, Children's Hospital, and Manager, National ARI Control Programme, Islamabad, Pakistan. a Khan MA, Rehman GN, Qazi SA. Control of acute respiratory infections in Pakistan: present status and future developments. Islamabad, Federal ARI Cell, Children's Hospital, Pakistan Insti- tute of Medical Sciences, 1990 (unpublished document): 10-21. Reprint No. 5731 this programme, the standardized ARI case manage- ment approach (7), recommended by WHO for the control of ARI, was adopted and modified to meet local needs.bc The modified strategy assesses the following clinical signs in children presenting with a cough or difficult breathing: fast breathing, indrawing of lower chest wall, and other specified danger signs. Anti- biotics are recommended for treatment of pneumo- nia, acute streptococcal pharyngitis, acute otitis media and mastoiditis, but not for treatment of acute upper respiratory infections such as coughs and colds. Commercial cough remedies containing inef- fective or harmful ingredients are discouraged. In 1990, ten ARI training units were established to train health professionals (physicians and paramedical staff); one of these is in the Children's Hospital, Pakistan Institute of Medical Sciences, Islamabad. Since health care professionals had not previously been trained in WHO's approach to ARI case man- agement, expensive antibiotics were often inappro- b National ARI Control Programme. Management of the young child with acute respiratory infection. Islamabad, Federal ARI Cell, 1990 (unpublished document). c National ARI Control Programme. Acute respiratory infections in children: case management in small hospitals in developing countries. A manual for doctors and other senior health workers. Islamabad, Federal ARI Cell, 1990 (unpublished document). Bulletin of the World Health Organization, 1996, 74 (5): 501-507 © World Health Organization 1996 501 S.A. Qazi et al. priately prescribed. From January 1990, physicians in the outpatient, emergency and inpatient depart- ments of the hospital adopted the modified standard ARI case management strategy. This article compares the inpatient ARI case fatality rates and extent of antimicrobial therapy in 1989 (before training) with those in 1990-92 (after training). Materials and methods The ARI training unit at the Children's Hospital, Islamabad, conducts 12-15 training courses every year with 12-15 participants in each course. Physi- cians, supervisors of health facilities, and para- medical staff are trained for 4-5 days in courses specially adapted for each category, which provide the knowledge and skills needed to assess, classify and treat children with ARI. Particular emphasis is placed on recognition and treatment of pneumonia. Prior to the introduction of this strategy pneumonia cases were diagnosed and admitted on the basis of relevant chest auscultation and radiographic findings. The hospital has 200 beds for medical, surgical and neonatal patients; 350-400 patients are seen daily from 08:00 to 14:00 in the outpatient depart- ment, and during the evening and night patients report to the emergency department. As there is no organized referral system in the community or with other health facilities, the majority of patients are self-referred. In the period January-March 1990, all regularly employed physicians in the hospital were trained in standard ARI case management. As the hospital is part of a teaching institute, new interns and resi- dents are also trained. In addition, the training staff supervised the trained physicians and held seminars and individual (one-to-one) sessions regularly. All children reporting to the hospital with ARI symptoms are sent by the registration desk to two specified clinics in the outpatient department, which are staffed by 3-4 trained physicians who attend to all under-5-year-olds with ARI and diarrhoea. In the outpatient and emergency departments, standard ARI case management guidelines for first- level facilities were followed.d From January 1990, the outpatient department recorded all demogra- phic, diagnostic and treatment information on patients with ARI in specifically designed ARI regis- ters, which replaced the abstract registers for patients' records. The latter included the patients' d See footnote b, p. 501. complaints, age, sex, nutritional status, diagnosis, treatment and outcome. The ARI register included the standard ARI terminology for assessment, classi- fication (no pneumonia, pneumonia, severe pneu- monia or very severe pneumonia), and management. Patients who had only a cough and cold (no pneumo- nia), sore throat, otitis media, acute pharyngitis, acute tonsillitis, or pneumonia (not severe) were sent home with oral medications and advice on home care (about feeding, giving fluids, clearing the nose, soothing the throat) and on when to return to the health facility. Patients diagnosed with pneumonia (not severe), in the presence of a cough and/or diffi- cult breathing and fast breathing (defined as ¢60 breaths per minute for under-2-month-olds, ¢50 for 2-11 months of age, and ¢'40 for 12-59 months of age), were prescribed oral antibiotics and sent home. The recommended antibiotics were oral cotrimoxa- zole for pneumonia (not severe) and otitis media and injectable benzathine penicillin for streptococcal sore throat, followed by oral amoxycillin as the alter- native drug for these three diagnoses.d Patients diag- nosed with severe pneumonia, with a cough and/or difficult breathing and lower chest wall indrawing, were referred for hospitalization. Age-specific danger signs were (1) for children <2 months of age: stopped feeding well, convulsions, abnormally sleepy or difficult to wake, grunting or groaning, wheezing, fever (>38°C), skin feels cold or shows cyanosis; (2) for children aged 2-59 months: not able to drink, convulsions, abnormally sleepy or difficult to wake up, stridor in a calm child, or clinically severe malnutrition.d Patients with a cough and/or difficult breathing and any of these age-specific dan- ger signs were suspected to be suffering from very severe disease (very severe pneumonia or septicae- mia or meningitis) and were referred for inpatient evaluation. If meningitis and septicaemia were excluded, these patients were treated for very severe pneumonia. Hospitalized severe and very severe pneumonia patients were managed according to the guidelines for standard ARI case manage- ment in hospitals.e These guidelines include use of antibiotics, oxygen and bronchodilator therapy as clinically indicated according to disease severity. In children aged 2-59 months, injectable benzyl penicillin or ampicillin was recommended for those classified as severe pneumonia and injectable chloramphenicol for those with very severe pneumo- nia. In children less than 2 months old a combination of injectable benzyl penicillin and gentamicin was recommended.de Oxygen was provided if the child had cyanosis, severe chest wall indrawing, respira- e See footnote c, p. 501. 502 WHO Bulletin OMS. Vol 74 1996 Management of acute respiratory infections in children in Pakistan Table 1: Characteristics of children with acute respiratory infections (ARI) presenting in the outpatient department of the Children's Hospital, Islamabad, 1989-1992 Patients and categories 1989 1990 1991 1992 No. of children seen 110386 102052 106949 80789 No. of under-5-year-olds 30333 (27.5)a 39107 (38.3) 36535 (34.2) 27701 (34.3) No. of under-5-year-olds with ARI 7986 (26.3) 10046 (25.7) 6548 (17.9) 5371 (19.4) Type of ARI in under-5-year-olds:b Not pneumonia, but other ARIc - 7006 (69.7) 4712 (72.0) 3926 (73.1) Pneumonia - 2818 (28.1) 1 633 (24.9) 1 093 (20.4) Severe pneumonia - 201 (2.0) 177 (2.7) 297 (5.5) Very severe pneumonia - 21 (0.2) 26 (0.4) 55 (1.0) No. of under-5-year-old ARI patients advised antibiotics 4358 (54.4) 3459 (34.4) 1 701 (26.0) 1 230 (22.9) and sent home No. of under-5-year-olds admitted and received antibiotics 160 (2.0) 259 (2.6) 238 (3.6) 393 (7.3) a Figures in parentheses are percentages. b In 1989, 6491 (81.3%) had acute upper respiratory infections and 1495 (18.7%) had acute lower respiratory infections. These are not WHO-recommended designations. c Other ARI such as wheezing, otitis media, streptococcal pharyngitis, etc. tory rate more than 20 above the age-specific cut-off point, extreme irritability, or blood oxygen satura- tion of <90%.' Patients were admitted from both the outpatient and emergency departments. ARI case fatality was defined as a patient admitted primarily with ARI who subsequently died in the hospital. The nutritional status of under-5-year-old inpatients was assessed by using the Gomez classification (8). The data for all analyses were abstracted from patients' records in the outpatient and inpatient departments. Until 1989 the hospital provided treatment free of charge to all patients, but in late 1990 some charges were introduced for patient registration, laboratory tests, and use of a hospital bed. However, medicines were provided free by the hospital phar- macy throughout the study period. Analysis and outcome variables. Analyses were carried out on information from the abstract regis- ters for all ARI patients in the outpatient and inpatient departments in 1989-92. The primary analysis was to investigate the impact of the standard ARI case management on ARI case fatalities among inpatients and on the extent of antibiotic use. Secondary analyses were conducted to assess the change over time in the severity of disease and the age structure and nutritional status of patients admitted with ARI. Stratification by analysis was done on children under 5 years. Significance levels were determined by the x2 test for linear trend in proportions from data over four years using EPINFO-6, a data management and analysis pack- age developed by the Centers for Disease Control and Prevention (CDC) and WHO. ' See footnote c, p. 501. Results Table 1 shows the characteristics of patients seen in the outpatient department during the period 1989- 92. The total number of patients decreased by 26.2% (from 110386 in 1989 to 80789 in 1992) over these four years, while the under-5-year-old ARI patients decreased by nearly 32.7% (7986 in 1989 to 5371 in 1992). Starting in 1990, pneumonia was classified according to the WHO-recommended ARI classifi- cation. No data were available for the categories of ARI in 1989. Together, all categories of pneumonia (pneumonia, severe pneumonia and very severe pneumonia) accounted for 30.3%, 28% and 26.9%, respectively, of all under-5-year-old ARI cases seen in the outpatient department from 1990 to 1992. Although the majority of patients with pneumonia were not severe cases, the proportion with either severe or very severe pneumonia increased from 2.2% in 1990 to 6.5% in 1992. The overall antibiotic use decreased by 46.6%, i.e. from 56.6% in 1989 to 30.2% in 1992 (P < 0.0001), among ARI pa- tients seen in the outpatient department. The major- ity of this reduction was observed in patients who previously would have been sent home on oral antibiotics. Table 2 shows that while total admissions (from outpatient and emergency departments) increased by 79% (from 3753 in 1989 to 6713 in 1992), the admissions for ARI more than doubled (from 776 in 1989 to 1673 in 1992). Distribution of all under-5- year-olds admitted with ARI also included children admitted with acute respiratory problems other than pneumonia. The proportion of very severe/severe pneumonia more than doubled from 1990 to 1992. Information about nutritional status was missing from the records of 24% of children in 1990-92. The WHO Bulletin OMS. Vol 74 1996 503 S.A. Qazi et al. Table 2: Number of medical and ARI admissions, by year, and characteristics of under-5-year-old ARI patients hospitalized at the Children's Hospital, Islamabad, 1989-1992 Characteristics 1989 1990 1991 1992 No. of medical admissionsa 3753 4275 6035 6713 No. of ARI admissionsa 776 (20.7)b 1103 (25.8) 1 432 (23.7) 1 673 (24.9) No. of under-5-year-olds with ARI, by age:c 1 month 41 70 109 106 2-11 months 421 576 780 995 12-59 months 223 337 381 356 Severity of pneumonia in under-5-year-olds: Severe and very severe pneumonia - 235 441 765 Pneumonia - 658 805 646 Nutritional status of under-5-year-olds with pneumonia:d Normal 261 410 561 729 Mild malnutrition 48 132 121 99 Moderate malnutrition 91 126 123 95 Severe malnutrition 86 102 141 138 a Includes children aged >5 years. b Figures in parentheses are percentages. c Includes patients with diagnoses other than pneumonia. d Gomez classification of nutritional assessment was used (8); information on nutritional status of several pneumonia patients was not available. distribution of the recorded categories of nutritional status remained similar from 1989 to 1992. Table 3 shows the overall and ARI-specific mor- tality stratified by age, severity of pneumonia, and nutritional status in hospitalized patients from 1989 to 1992. The overall case fatality rate (CFR) for all admissions decreased by 28% from 8.7% in 1989 to 6.2% in 1992 (P < 0.0001), while that for ARI decreased by 50.5% from 9.9% in 1989 to 4.9% in 1992 (P < 0.0001). CFR remained high for neonates. There was a decrease in case fatality among 1-11- month-old children (P < 0.0001) and those aged 12-59 months (P = 0.011). CFR decreased in chil- dren with very severe/severe pneumonia from 34.9% in 1990 to 7.8% in 1992 (P < 0.0001), and also fol- lowing admission for non-severe pneumonia from 5.2% to 2.6% (P = 0.02). There was a significant reduction in CFR only in those children recorded as severely undernourished (P = 0.01). Discussion Our data show two important findings: a remarkable decrease in antibiotic use over the study period, and a significant decrease in ARI case fatality among the Table 3: Number of deaths and case fatality rates among medical and ARI admissions and their distribution by age group, severity of pneumonia, and nutritional status, 1989-1992 Characteristics 1989 1990 1991 1992 No. of deaths among medical admissionsa 328 (8.7)b 435 (10.2) 491 (8.1) 419 (6.2) No. of deaths among ARI admissionsa 76 (9.9) 137 (12.4) 113 (7.9) 82 (4.9) No. of deaths among under-5-year-olds with ARI, by age:c >1 month 10 (24.4) 22 (31.4) 41 (37.6) 29 (27.4) 2-11 months 52 (12.4) 82 (14.2) 54 (6.9) 36 (3.6) 12-59 months 10 (4.5) 26 (7.7) 9 (2.4) 9 (2.5) Severity of pneumonia in under-5-year-olds: Severe and very severe pneumonia - 82 (34.9) 74 (16.8) 60 (7.8) Pneumonia - 34 (5.2) 35 (4-4) 17 (2.3) Nutritional status of under-5-year-olds with pneumonia:d Normal 9 (3.5) 11 (2.7) 15 (2.7) 14 (1.9) Mild malnutrition 4 (8.3) 17 (12.9) 5 (4.1) 9 (9.1) Moderate malnutrition 17 (18.7) 28 (22.2) 22 (17.9) 13 (13.7) Severe malnutrition 23 (26.7) 35 (34.3) 27 (19.2) 24 (17.4) a Includes children aged >5 years. b Figures in parentheses are percentages; they indicate the case fatality rates. c Includes patients with diagnoses other than pneumonia. d Gomez classification of nutritional assessment was used (8). WHO Bulletin OMS. Vol 74 1996504 Management of acute respiratory infections in children in Pakistan hospitalized children. The first finding demonstrates that standard ARI case management can reduce inappropriate antibiotic therapy. This reduction is justified because most acute upper respiratory infections are viral in etiology (7) and only a small proportion are bacterial requiring antibiotics, e.g., streptococcal pharyngitis and otitis media. In a com- munity study carried out in four villages around Islamabad in 1991, 80% of the ARI patients seen were successfully managed without antibiotics (9). In our outpatient data, 80% of ARI cases over a period of four years were acute upper respiratory infections, the majority of which did not need antibiotics. The overuse of antibiotics for ARI therapy in 1989 and 1990 in our hospital was similar to that reported elsewhere (10-12). Indiscriminate use of antibiotics has attracted much more attention recently because it has led to increasing resistance of, for example, Haemophilus influenzae and Streptococcus pneu- moniae, the two commonest organisms responsible for pneumonia in Pakistan and elsewhere (13, 14). The reduction in use of antibiotics for ARI therapy was sustained over four years in our study. Training of all physicians in standard ARI case management, display of these management charts in the outpa- tient, emergency and inpatient departments, close supervision, monitoring and support were most probably responsible for the sustained reduction of antibiotic use. Similar results have been reported elsewhere.9 The standard ARI case management approach also reduces health care costs. Direct savings can be achieved by eliminating the use of antibiotics and cough syrups when these drugs are not indicated for treatment of acute upper respiratory infections. In 1992, our health facility saved nearly Rs 100000 (US$ 3077) through improved prescribing practices Although our hospital's prescribing practices may not be representative of those in other health facili- ties in the country, data from three large metropoli- tan cities in different parts of Pakistan suggest that inappropriate drug use is widespread. A study on prescribing practices in Karachi reported that 53% of contacts between patients and general physicians resulted in an antibiotic prescription (15). A recent study of general physicians' prescribing practices for ARI conducted in Multan showed an average of 3.4 prescribed medicines (antibiotic, cough remedy, antipyretic, anti-allergic or vitamin preparation) for each ARI episode, and that antibiotics were pre- scribed for 72% of acute upper respiratory infections 9 WHO Programme for Control of Acute Respiratory Infec- tions. Interim programme report 1992. Unpublished report WHO/ ARI/93.25, 1993: 22-32. (16). Based on the cost savings at our hospital, a conservative estimate of savings can be made for the whole country. If there are 110 million episodes of ARI annually among the 22.5 million children in Pakistan (2) and only 25% of these were seen and treated by health professionals, the cost of prescrib- ing antibiotics (cotrimoxazole, which is the cheapest) to 55% of children (pre-intervention estimate) would be Rs 126.2 million (US$ 3.9 million) annu- ally, or about 10% of the allocation to Pakistan's public health budget in 1994-95 (Rs 1150 million or US$ 37.1 million). Appropriate ARI case manage- ment could save Rs 39.8 million (US$ 1.2 million) from the budget, and a reduction in indiscriminate use of commercial cough remedies could save Rs 275 million (US$ 8.5 million) annually. The ARI case fatality rate in 1989-90 at our hospital (9.9%) was similar to that in other large teaching hospitals in Pakistan: 9.4% from the National Institute of Child Health in Karachi,h and 10% from Ganga Ram Hospital in Lahore (17). These rates are similar to those reported from other countries: 15% CFR for severe pneumonia cases from three hospitals in Papua New Guinea (18) and 10% CFR for Kenyan children admitted with acute lower respiratory infections (ALRI) at the Kenyatta General Hospital (19). With the introduction of stan- dard ARI case management, the reduction in CFR at our hospital over a period of four years (9.9% to 4.9%) is similar to that reported from Napoleon Franco Pareja Children's Hospital in Cartagena, Colombia, (from >13% in 1988 to 4% in 1991 in children age 2 months to 4 years, following introduc- tion of WHO's standard ARI case management in 1988).9 This reduction in ARI case fatality rates con- firms the effectiveness of the ARI case management strategy, which provides very specific guidelines on the management of pneumonia cases and the use of antibiotics and oxygen. Hypoxaemia and ALRI- related deaths appear to be associated (17), which suggests that early detection and effective manage- ment of hypoxaemia is a key component of ALRI management. We have no data on the oxygen pro- vided to our pneumonia patients, but prior to adopt- ing the standard ARI case management guidelines no standard criteria were followed for oxygen sup- plementation at our hospital. It is quite possible that the early detection of hypoxaemia and provision of hKhan MA, Qazi SA, Rehman GN. Control of acute respiratory infections in Pakistan: present status and further developments. In: Proceedings of a Workshop on the National ARI Control Programme, 13-14 December 1989. Islamabad, Federal ARI Cell, 1990. WHO Bulletin OMS. Vol 74 1996 505 S.A. Qazi et al. oxygen contributed to the lower mortality in admit- ted patients, particularly those suffering from very severe/severe pneumonia. A closer examination of the data shows that, over the four-year study period, there was more than a twofold increase in ARI admissions but the actual number of deaths remained relatively constant. It is possible that severe pneumonia cases were being overdiagnosed, either because of more sensitive diagnostic criteria or because physicians were afraid of missing a child with severe pneumonia, thus resulting in a spurious drop in case fatality rates. Many children admitted during 1990-92 would not have been admitted if the 1989 criteria, based on clinical auscultation and radiological diagnosis, had been followed. The possibility that a large number of nonsevere pneumonia patients were misclassified and admitted (although admission was not recom- mended) cannot be ruled out. It is also possible that some of the patients recorded as pneumonia in the inpatient registers (the source of this information) were in fact severe pneumonia. ARI admissions also increased because a clinical trial, conducted between November 1991 and April 1992 to study the efficacy of oral cotrimoxazole versus oral amoxycillin, en- rolled 350 patients nearly equally divided between nonsevere pneumonia and severe pneumonia. Finally, it should be noted that while non-ARI admissions increased nearly twofold during the study period, the number of deaths remained fairly constant. Our hospital facility is relatively new and provides good tertiary care, so an increasing number of very sick children were being seen at the hospital. On the basis of our results it is very difficult to assess the separate effects of ARI standard case management and overdiagnosis of severe pneumo- nia cases on case fatality rates. We think both factors contributed to this reduction. If we had used only auscultation and radiography and not admitted the children labelled as severe pneumonia, we have no way of knowing how many would have succumbed without injectable antibiotics. We do know, how- ever, from previous community studies in Pakistan and elsewhere that using standard ARI case man- agement saves lives (1, 20). Two other points should be mentioned. First, the lack of a decline in neonatal case fatality is striking. Our neonatal intensive care unit routinely admits very seriously sick children, and surveillance has provided limited information on (i) co-existence of neonatal septicaemia with pneumonia resulting in a misclassification of deaths due to ARI; and (ii) the role of prematurity and small-for-gestational age on mortality. Second, the total outpatient visits declined from 110386 in 1989 to 80789 in 1992. This may have occurred for two reasons. User charges, which were introduced in 1990, often result in reduced use of health services (21,22). Since ARI case management eliminated unnecessary therapies, some "sick" chil- dren could have been taken elsewhere for other procedures and to receive antibiotics. A meta-analysis of six published community- based ARI standard case management intervention trials from various countries showed an overall reduction of 20% in infant mortality and an overall reduction of 25% in under-five mortality (20). Using two simple danger signs, lower chest wall indrawing and fast breathing, community health workers have been able to identify pneumonia. Lessons learned from the successful use of this strategy by community health workers can be utilized in the facilities providing first-level health care. By implementing the ARI control programme, both ARI mortality and overall childhood mortality can be reduced, and the cost of health care decreased by eliminating unnecessary antibiotic use, which would also reduce microbial resistance. We there- fore recommend the adoption of standard ARI case management for all patients seen in hospital out- patient departments and peripheral health facilities throughout Pakistan. Acknowledgements We are indebted to the following for reviewing the manu- script and for their suggestions: Dr Patricia Hibberd, Mr Jonathan Simon, Ms Claudia Johnson, Dr Richard Cash and Mr Jonathan Harrington (Applied Diarrheal Disease Project, Harvard Institute of International Development, Harvard University, Cambridge, MA, USA); Dr David Robinson, Dr Antonio Pio and Dr Sumaira Abu Bakr (WHO, Geneva, Switzerland); and Dr Harry Campbell (Child Health Dialogue, AHRTAG, London, England), Professor Andrew Tomkins (Institute of Child Health, London, England); and Dr Walter Straus (Department of Gastroenterology, University of North Carolina Hospital, North Carolina, USA). We also thank Professor K.A. Abbas, Dr S. Qureshi, Dr M. Hassan, Dr M. Jamal, Dr M. Azam, Dr M. Khalid and Dr J. lqbal, consultants at the Children's Hospital, under whose care the patients were admitted. Resume Prise en charge standardisee des infections respiratoires aigues dans un h8pital pediatrique du Pakistan: repercussions sur l'utilisation des antibiotiques et les taux de letalite Les infections respiratoires aigues (IRA) sont une des causes majeures de morbidite et de mortalite WHO Bulletin OMS. Vol 74 1996506 Management of acute respiratory infections in children in Pakistan infantiles au Pakistan. Le programme national de lutte contre les IRA a et6 lanc6 en 1989 dans le but de r6duire la mortalit6 due a la pneumonie et de rationaliser l'utilisation des antibiotiques dans la prise en charge des malades atteints d'IRA. Les directives de l'OMS pour la prise en charge des cas d'IRA ont 6te adopt6es. Le personnel m6dical du Children's Hospital d'Islamabad a ete form6 a ces directives au d6but de 1990, et le nouveau person- nel a 6te form6 lors de sessions compl6mentaires. Les donnees sur les malades vus en am- bulatoire ont ete fournies par des registres sp6ciaux pour les IRA tenus par le service des consultations externes depuis janvier 1990. Les donn6es sur les malades hospitalises pour IRA ont 6t6 fournies par les registres de l'hopital. Pendant la p6riode de quatre ans consid6r6e (1989 a 1992), le taux d'utilisation des antibiotiques dans le service des consultations externes est tombe de 54,6% a 22,9% (p < 0,0001). Le taux de letalit6 chez les enfants hospitalis6s pour IRA est tomb6 de 9,9% a 4,9% (p < 0,0001) et le taux de l6talit6 global, de 8,7% a 6,2%. Nos resultats montrent que la prise en charge standardis6e des cas d'IRA dans un 6tablissement de soins tertiaire a reduit a la fois l'utilisation des antibiotiques et les d6penses en medicaments. Bien que les criteres de prise en charge des cas d'IRA, qui sont plus sensibles que les criteres clas- siques reposant sur l'auscultation et la radiographie, aient entrainM un plus grand nombre d'hospita- lisations, nous estimons que cette strat6gie a con- tribue a r6duire sensiblement le taux de l6talit6 dO a ces infections. References 1. Khan AJ et al. Acute respiratory infections in children: a case management intervention in Abbottabad Dis- trict, Pakistan. Bulletin of the World Health Organiza- tion, 1990, 68: 577-585. 2. Townsend EH, Radebaugh JF. Prevention of compli- cations of respiratory illness in pediatric practice: a double-blind study. New England journal of medicine, 1962, 226: 683-689. 3. Ackerman BD. Treatment of undifferentiated respira- tory infections in infants. Clinical pediatrics, 1968, 7: 391-395. 4. Hardy LM, Traisman HS. Antibiotics and chemo- therapeutic agents in the treatment of uncomplicated respiratory infections in children. Journal ofpediatrics, 1956, 48: 146-156. 5. Lexomboon U et al. Evaluation of orally administered antibiotics for treatment of upper respiratory infections in Thai children. Joumal of pediatrics, 1971, 78: 772- 778. 6. Taylor B et al. Amoxycillin and cotrimoxazole in presumed viral respiratory infections of childhood: placebo-controlled trial. British medical journal, 1977, 2: 552-554. 7. Acute respiratory infections in children: case manage- ment in small hospitals in developing countries. A manual for doctors and other senior health workers. Geneva, World Health Organization, 1990. 8. Reddy V. Protein-energy malnutrition. In: Stanfield P, ed. Diseases of children in the subtropics and tropics. London, Edward Arnold, 1991: 335-337. 9. Khan MA et al. A community study of the appli- cation of WHO ARI management guidelines in Pakistan. Annals of tropical paediatrics, 1993, 13: 73-78. 10. Abbott GD, Fergusson DM, Horwood LJ. General practitioner prescribing practices for childhood respi- ratory infection. New Zealand medical journal, 1982, 95: 185-188. 11. Hossain MM, Glass RI, Khan MR. Antibiotic use in rural community in Bangladesh. International journal of epidemiology, 1982, 11: 402-405. 12. Maitai CK, Watkins WM. A survey of out-patient prescriptions dispensed in Kenyatta National Hospital. East African medical journal, 1980, 58: 641-645. 13. Mastro TD et al. Antimicrobial resistance of pneu- mococci in children with acute lower respiratory tract infections in Pakistan. Lancet, 1991, 337, 156- 159. 14. Weinberg GA et al. and the BOSTID Haemophilus Susceptibility Study Group. Antimicrobial suscepti- bility patterns of Haemophilus isolates from children in eleven developing nations. Bulletin of the World Health Organization, 1990, 68, 179-184. 15. Nizami SO, Khan IA, Bhutta ZA. Pediatric prescrib- ing practices in Karachi, Pakistan. Journal of Pakistan Medical Association (in press). 16. Applied Diarrhoeal Disease Project. Research highlights. In: Annual report 1994. Cambridge, MA, Harvard Institute of International Development, 1994: 1-16. 17. Mannan J. Significance and evaluation of gastric lavage diagnosis of pathogens in pneumonia. Paki- stan paediatric joumal, 1985, 9: 36-40. 18. Shann F, Barker J, Poor P. Clinical signs that predict death in children with severe pneumonia. Pediatric infectious diseases journal, 1989, 8: 852- 855. 19. Onyango FE et al. Hypoxemia in young Kenyan chil- dren with acute respiratory infections. British medical journal, 1993, 306: 612-615. 20. Sazawal S, Black R. Meta-analysis of inter- vention trials of case management of pneumonia in community settings. Lancet, 1992, 340: 528- 533. 21. Kanji N. Charging for drugs in Africa: UNICEF's "Bamako Initiative". Health policy and planning, 1989, 4: 110-120. 22. Creese AL. User charges for health care: a review of recent experience. Health policy and planning, 1991, 6: 309-319. WHO Bulletin OMS. Vol 74 1996 507