Effectiveness of clinical guidelines for the presumptive treatment of streptococcal pharyngitis in Egyptian children.

BACKGROUND: Primary prevention of acute rheumatic fever requires antibiotic treatment of acute streptococcal pharyngitis. In developing countries, clinicians must rely on clinical guidelines for presumptive treatment of streptococcal pharyngitis since bacterial culture and rapid diagnostic tests are not feasible. We evaluated the WHO Acute Respiratory Infection guideline in a large urban paediatric clinic in Egypt. METHODS: Children between 2 and 13 years of age who had a sore throat and pharyngeal erythema were enrolled in the study. Clinical, historical, and demographic information was recorded and a throat culture for group A beta-haemolytic streptococci was done. Sensitivity (% of true-positive throat cultures) and specificity (% of true-negative throat cultures) were calculated for each clinical feature. The effect of various guidelines on correct presumptive treatment for throat-culture status was calculated. FINDINGS: Of 451 children with pharyngitis, 107 (24%) had group A beta-haemolytic streptococci on throat culture. A purulent exudate was seen in 22% (99/450) of these children and this sign was 31% sensitive and 81% specific for a positive culture. The WHO Acute Respiratory Infections (ARI) guidelines, which suggest treatment for pharyngeal exudate plus enlarged and tender cervical node, were 12% sensitive and 94% specific; 13/107 children with a positive throat culture would correctly receive antibiotics and 323/344 with a negative throat culture would, correctly, not receive antibiotics. Based on our data we propose a modified guideline whereby exudate or large cervical nodes would indicate antibiotic treatment, and this guideline would be 84% sensitive and 40% specific; 90/107 children with a positive throat culture would correctly receive antibiotics and 138/344 with a negative throat culture would, correctly, not receive antibiotics. INTERPRETATION: The WHO ARI clinical guideline has a high specificity but low sensitivity that limits the unnecessary use of antibiotics, but does not treat 88% of children with a positive streptococcal throat culture who are at risk of acute rheumatic fever. A modified guideline may be more useful in this population. Prospective studies of treatment guidelines from many regions are needed to assess their use since the frequency of pharyngitis varies.

PIP: In developing country settings without access to bacterial culture and rapid diagnostic tests, the prevention of acute rheumatic fever depends on clinicians' presumptive treatment of streptococcal pharyngitis. This study evaluated the effectiveness of World Health Organization (WHO) acute respiratory infection guidelines in a large pediatric clinic (Abu Reesh Children's Hospital) in Cairo, Egypt. 451 children 2-13 years of age with sore throat and pharyngeal erythema were enrolled, 107 (24%) of whom had group A beta-hemolytic streptococci on throat culture. Purulent exudate, present in 99 (22%) of these children, was 31% sensitive and 81% specific for a positive culture. The WHO guidelines, which recommend treatment for pharyngeal exudate plus enlarged and tender cervical node, were 12% sensitive and 94% specific. Based on these guidelines, 13 of 107 children with a positive throat culture would correctly receive antibiotics and 323 of 344 with a negative culture would not receive antibiotics. A modified guideline in which exudate or large cervical nodes would indicate antibiotic treatment would be 84% sensitive and 40% specific. With this modification, 90 of 107 children with a positive throat culture would correctly receive antibiotics and 138 out of 344 with a negative culture would not receive treatment. However, additional prospective studies from other regions of Egypt are necessary before modified guidelines are implemented.

A national programme for the control of acute respiratory infections.

In 1989 Egypt set up a programme to combat acute respiratory infections among children aged under 5 years. The present article indicates what has been achieved and the lessons that have been learnt.

PIP: Pediatricians, epidemiologists, and public health administrators largely agreed in the 1980s that acute respiratory infections (ARI) were the largest neglected threat to child survival in developing countries. While some scientists believed that large-scale studies should be conducted to develop control methods, others believed that prevention through immunization was the only viable approach. In response to prevailing relatively high levels of morbidity and mortality associated with ARI among children under age 5 years, Egypt in 1989 established a national program to fight ARI among such young children. The main control strategy of the program was standard case management for early detection and therapy using health care providers who diagnose and treat children with ARI, especially pneumonia. What has been achieved and lessons learned are described. Sections discuss the decentralization of planning and implementation, quality and quantity in training activities, communication activities, the availability of services, maintenance and repair of equipment, monitoring delivery of case management, selective surveillance and research, and the involvement of leading professionals.

Resistance patterns of Streptococcus pneumoniae and Haemophilus influenzae isolates recovered in Egypt from children with pneumonia. The Antimicrobial Resistance Surveillance Study Group.

Treatment of childhood pneumonia in developing countries requires knowledge of susceptibility patterns for Streptococcus pneumoniae and Haemophilus influenzae. Between October 1991 and April 1993, a surveillance survey of antimicrobial resistance was performed at two fever hospitals in Egypt; nasopharyngeal swab and blood specimens obtained from 1,635 children with pneumonia were cultured for these organisms. Susceptibility testing of these organisms was performed. At least one of these organisms was isolated from nasopharyngeal swab specimens from 73% of the children; 3.7% of blood cultures were positive. For S. pneumoniae strains, 70.9% of nasopharyngeal isolates were calculated to be susceptible to penicillin vs. 77.6% of blood isolates; the percentages of isolates susceptible to co-trimoxazole were 73.0% and 75.0%, respectively. For H. influenzae strains, 93.0% of nasopharyngeal isolates were calculated to be susceptible to ampicillin vs. 100% of blood isolates; the percentages of isolates susceptible to co-trimoxazole were 84.9% and 100%, respectively. Although most S. pneumoniae and H. influenzae strains associated with childhood pneumonia in Cairo were susceptible to penicillins and co-trimoxazole, antimicrobial resistance did not occur.

Estimating the completeness of under-5 death registration in Egypt.

To evaluate the completeness of registration of infant and child deaths in Egypt, reinterviews were conducted with families who had reported a death of a child under age 5 in the five years before the survey for two national surveys recently conducted in Egypt: the United Nations PAPCHILD survey of 1990-1991 and the Egyptian Demographic and Health Survey (EDHS) of 1992. The survey instrument included questions regarding notification of the death at the local health bureau. If the family said the death had been notified, separate employees searched the health bureau records for the registration. Overall 57% of infant deaths were reported as notified and 68% of those death reports were found; the corresponding figures for child death were 89% and 74%. Using the percentage reported as notified as an estimate for completeness of registration, we adjusted upward the national infant and child mortality rates from registration data, giving values of 73 per 1,000 for infant mortality and 99 for 5qo for the period 1987-1990. These values are approximately 20% above the corresponding direct estimates from the PAPCHILD and EDHS surveys.

Field trial of oxygen concentrators in upper Egypt.

Technical problems in developing countries often require more than just technological solutions. Many small hospitals in rural areas are without a reliable oxygen supply; small oxygen concentrators offer a solution, but simply sending out machines is ineffective. This account details the setting up and first year's operation of a project to test oxygen concentrations in a developing country. A co-ordinated strategy has been developed to include machines, supplies, education, training, and feedback. Initial results are encouraging, and support the idea that suitably installed and maintained concentrations can have a valuable role in bringing oxygen therapy to patients and hospitals in countries which have so far been denied it.

Acute respiratory infections: sentinel survey in Egypt.

A survey of 115 health centres in five governorates provided morbidity figures for acute respiratory infections (ARI) from 75,789 records. The findings give an estimate of antibiotic requirements, including dosage forms for young children, and help evaluate the performance of individual health centres and the entire ARI programme.

PIP: To obtain accurate data on the incidence and clinical profile of acute respiratory infections (ARI) in Egyptian infants and children aged 4 years and under, 115 randomly selected health centers in five of the country's 26 governorates participated in an ARI Case Registration System. Each month during 1993, specially trained physicians assigned to the project submitted a case diagnosis form. The survey yielded a total of 75,789 records representing seven categories: very severe disease (0.9%); severe pneumonia (2.9%); pneumonia (9.9%); otitis media (10.7%); pharyngitis (22.8%); asthma (4.8%); and cough, cold, and bronchitis (48.1%). 4.1% of patients were under 2 months old, 29.9% were 2-11 months of age, and 66.0% were 1-4 years old. Pneumonia cases peaked in January and fell off in June. When the reliability of disease symptoms as reported by caretakers was compared to the corresponding clinical signs observed by physicians, specificity was in the 95-97% range, but sensitivity was extremely low (e.g., 48-59% for rapid breathing). This analysis enabled health centers to calculate the annual drug requirements for use in ARI. For example, the finding that 34% of ARI cases involve children under 12 months of age led to a decision to ensure that one-third of the amoxycillin ordered for use in the clinics is of the 125 mg strength.

Monovalent latex agglutination reagents for the diagnosis of nonmeningitic pneumococcal infection.

The pneumococcus is a leading cause of serious bacterial infection worldwide. Given the difficulties with available assays for the diagnosis of invasive nonmeningitic pneumococcal infection, we evaluated monovalent slide latex agglutination reagents among patients with blood culture-confirmed pneumococcal infection and control patients in Baltimore, Maryland, USA; São Paulo, Brazil; and Cairo, Egypt. Among 50 patients with invasive nonmeningitic pneumococcal infection, 23 had a positive urine test for a sensitivity of 46% (95% confidence intervals of 32% and 61%). Among 39 healthy children, 36 had a negative assay, for a specificity of 92% (95% confidence intervals of 78% and 98%). Among 80 children with pneumonia without a positive blood culture for Streptococcus pneumoniae, the specificity was 88% (95% confidence intervals of 78% and 94%). Although the assay was fairly specific, the positive predictive value using optimistic assumptions was only 73%-83%. This study suggests that this assay has a sensitivity and positive predictive value that may limit its value in some settings.

Maternal reporting of acute respiratory infection in Egypt.

BACKGROUND: Acute respiratory infection (ARI) is a major cause of childhood morbidity and mortality in developing countries. Community surveys are used to determine the proportion of children with ARI for whom care is sought by questioning mothers about the signs and symptoms of illness episodes. The validity of this approach has been studied infrequently. METHODS: We evaluated maternal reporting of signs and symptoms 2 and 4 weeks after diagnosis among 271 Egyptian children < 5 years old. Children with ARI were evaluated by physical examination, chest radiography, and pulse oximetry, and were alternately assigned for a maternal interview about the episode 14 or 28 days later. RESULTS: For radiographically-defined acute lower respiratory infection (ALRI), the sensitivity of several symptoms for combined open- and close-ended questions was relatively high: nahagan (deep or rapid breathing) (80%), nafas sarie (fast breathing) (66%), and kharfasha (coarse breath sounds) (63%). The specificity of these terms was 50-68%. The specificity was inversely related to the follow-up time. No term provided both a sensitivity and specificity of > 50% at day 28 across the radiographically, clinically- and pulse oximetry-based definitions of ALRI. Spontaneously mentioned karshet nafas (difficult or rapid breathing) at 14 days had a specificity and sensitivity for radiographic ALRI of 87% and 41%, respectively, suggesting that this term is a good choice for community surveys. CONCLUSIONS: Maternal reporting of ARI symptoms is non-specific 2 and 4 weeks after diagnosis but may be useful for monitoring trends in the proportion of children with pneumonia who receive medical care. To maximize specificity, ARI programmes should generally use a recall period of 2 weeks.

PIP: Acute respiratory infection (ARI) is a major cause of childhood morbidity and mortality in developing countries. Community surveys are used to determine the proportion of children with ARI for whom care is sought by questioning mothers about the signs and symptoms of illness episodes. The validity of this approach has been studied infrequently. Maternal reporting of signs and symptoms was evaluated 2 and 4 weeks after diagnosis among 271 Egyptian children 5 years old selected from a prospective study of the signs and symptoms of pneumonia in Ismailia and Assiut. Children with ARI were evaluated by physical examination, chest radiography, and pulse oximetry, and were alternately assigned for a maternal interview about the episode 14 or 28 days later. Nahagan (deep or rapid breathing), yenet (grunting), and karshet nafas (difficult or rapid breathing) were the most frequently mentioned terms reported by 49%, 28%, and 31% of the mothers, respectively. For radiographically-defined acute lower respiratory infection (ALRI), the sensitivity of several symptoms for combined open- and close-ended questions was relatively high: nahagan (80%), nafas sarie (fast breathing) (66%), and kharfasha (coarse breath sounds) (63%). The specificity of these terms was 50-68%. In general, the sensitivity was greater for 14- or 28-day recall than day 0, but the specificity was inversely related to the follow-up time. No term provided both a sensitivity and specificity of 50% at day 28 across the radiographically, clinically- and pulse oximetry-based definitions of ALRI. Spontaneously mentioned karshet nafas (difficult or rapid breathing) at 14 days had a specificity and sensitivity for radiographic ALRI of 87% and 41%, respectively, suggesting that this term is a good choice for community surveys. Maternal reporting of ARI symptoms is non-specific 2 and 4 weeks after diagnosis but may be useful for monitoring trends in the proportion of children with pneumonia who receive medical care. To maximize specificity, ARI programs should generally use a recall period of 2 weeks.

Caretaker compliance with different antibiotic formulations for treatment of childhood pneumonia.

To determine the effect of antibiotic formulations on compliance, 400 children, aged 2 months to 5 years, with a presumptive diagnosis of pneumonia, were randomly assigned to receive one of the following formulations of cotrimoxazole: 1. syrup accompanied by a 10-ml measuring cup; 2. syrup accompanied by a 5-ml measuring spoon; 3. tablets; 4. single-dose sachets of antibiotic powder. A research assistant visited the childs' home on the fourth day of therapy, asked the care-giver about compliance, and observed the care-giver prepare a dose of the medication. The remaining amount of medicine was measured, and when possible (n = 151), a urine specimen was tested for the presence of sulphamethoxazole. All of the care-givers reported giving at least one dose on the first day of therapy. By the fourth day, 82 per cent of those receiving syrup were still taking their medication compared to 71 and 55 per cent of those receiving sachets or tablets, respectively (P < 0.01). Of those who received syrup accompanied by a spoon, 38 per cent under-dosed the medicine by at least 30 per cent. Overall, compliance was highly correlated with the care-giver's report of difficulty in administering the medication. Additional research is needed to understand the obstacles encountered by care-givers in administering sachets and tablets. Meanwhile, the use of antibiotic syrup, accompanied by an appropriately sized measuring cup, appears to offer the greatest probability of medication compliance in the treatment of Egyptian children with pneumonia.

Hemophilus influenzae and Streptococcus pneumoniae as causative agents of pneumonia in Egyptian preschool children: analysis and serotyping of Hemophilus isolates from hospital patients in Cairo, 1991-93.

The Hemophilus influenzae blood culture and nasopharyngeal isolates, collected during a limited Surveillance of Antimicrobial Resistance in 1991-1993 from 1,635 Egyptian children under 5 years of age, presenting with pneumonia to Embaba and Abbassia Fever Hospitals, were serotyped. The 8 blood culture isolates confirmed H. influenzae to be responsible for 0.5% of the cases of pneumonia, versus 50 Streptococcus pneumoniae blood culture isolates from the same population that confirmed it responsible for 3.1% of the cases. The invasive Hemophilus strains were exclusively isolated from infants below 1 year, from one hospital (Embaba), on one winter season (January to March, 1992). On serotyping, 50% of the blood culture isolates were found to be non-b by latex agglutination. Some 297 nasopharyngeal isolates from cases of pneumonia were also serotyped and 45% were found to be non-b, thus confirming the invasive strains findings. Furthermore, the typing results from ARI-free controls nasopharyngeal isolates--though limited--were consistent with the findings and showed a 43% proportion of non-b. These findings put a question mark on the benefit of a large scale use of the available H. influenzae type b polysaccharide and conjugated vaccines in Egypt. But before interpreting out data in terms of vaccine needs, more specifically designed epidemiological studies need to be conducted to assess the role of H. influenzae as a pathogen in Egypt.

Developing strategies to encourage appropriate care-seeking for children with acute respiratory infections: an example from Egypt.

Determinants of care-seeking and patterns of referral for acute respiratory infections (ARIs) in children were studied in two communities in Ismailia, Egypt. A video was used to assess mothers' recognition and interpretation of clinical signs of serious illness. Mothers were questioned about which of the locally available provider options they would choose for four different locally-defined ARI illnesses; they were also read brief descriptions of hypothetical cases, and asked how they would recommend treating children in those situations. These results were compared with reported care-seeking practices during past ARI episodes that occurred in their own children. The results indicate that mothers generally recognize rapid or difficult breathing, but do not use the recognition to take appropriate actions. The data suggest that a substantial proportion of children in the study area, who are perceived to have severe respiratory illnesses, may not be brought to the government health facilities for treatment. Implications of the findings for the training policies and strategies of the Egyptian national ARI program are considered.

PIP: In Ismailia, Egypt, interviews with key informants and mothers of young children and presentations of a video were conducted in a suburban area with access to health facilities and in a rural village 3 km from the nearest public health clinic. The researchers wanted to assess mothers' recognition and interpretation of clinical signs of serious illness and to determine their preference of provider options by different locally defined acute respiratory infection (ARI) illnesses. Mothers heard descriptions of hypothetical cases and then reported how they would treat the children in the case scenarios. The researchers compared results from these data collection practices with the actual care-seeking practices of mothers during past childhood ARI episodes. When shown a video of 20 children with either no respiratory symptoms or mild to severe symptoms, mothers were able to correctly identify children who did not have an ARI in 56% of cases. With no prompting, they were able to correctly identify those with fast respiratory rates in 65% of cases. Mothers took their children with an ARI outside of the home for treatment in 22 of the past 30 ARI episodes. The most common reasons for taking children to a private physician in cases of serious illness rather than to a physician at a government health clinic were confidence in physician's ability and convenience, especially in hours of operation. Mothers tended to treat children with a runny nose and cough at home, regardless of age. Most did not consider fast and irregular breathing as a reason to seek treatment outside of the home. These results suggest that the National ARI Program should sponsor a media campaign to promote confidence in government providers to treat serious ARIs, call for a change in clinic schedules to increase access to care, and assume educational and managerial actions to assure quality of care and the availability of necessary drugs.

Oral versus nebulized albuterol in the management of bronchiolitis in Egypt.

The efficacy of albuterol in the management of infants with first-time wheezing is controversial. This double-blind, placebo-controlled trial involved 128 Egyptian infants with first-time wheezing (mean age 5.9 months) who were randomly separated into four equivalent therapy groups: nebulized albuterol, nebulized saline solution, orally administered albuterol, and orally administered placebo. The nebulized therapy groups received two treatments 30 minutes apart; the oral therapy groups received one treatment. In addition, 41 infants with recurrent wheezing comprised an open-label control group who received two albuterol nebulizations. Chest radiographs, leukocyte counts, blood culture specimens, and nasal aspirate for viral antigen detection were obtained to confirm the clinical diagnosis of viral bronchiolitis. Respiratory and heart rates, clinical score, and oxygen saturation were recorded at baseline and 30 and 60 minutes after treatment. There was no difference among the four randomized groups in any of the outcomes except for an increase in heart rate of 10 beats/min in the nebulized albuterol group compared with a decrease in heart rate of 8 beats/min in the oral placebo group. No differences in the need for additional treatment were observed. Nebulized albuterol significantly improved the clinical score and reduced the respiratory rate of those with recurrent wheezing relative to those in the randomized groups. We conclude that nebulized albuterol is effective in the treatment of infants with recurrent wheezing, but there was no demonstrable efficacy of orally administered or nebulized albuterol in relieving the respiratory distress of infants with bronchiolitis in Egypt.

Caretaker recognition of respiratory signs in children: correlation with physical examination findings, x-ray diagnosis and pulse oximetry.

Caretaker recognition of clinical utility of respiratory signs and symptoms in the prediction of pneumonia was examined in a prospective study of infants and children in four cities in Egypt. In all 688 children aged 2 months-5 years presenting with a history and/or physical examination findings of cough and difficult or fast breathing were recruited from out-patient health facilities. The validity of caretaker terms was determined using paediatrician observation of standard respiratory signs and symptoms, x-ray diagnosis and pulse oximetry as standards. The sensitivity of 'nahagan' (Egyptian Arabic for fast breathing) for identifying elevated respiratory rate was 78% +/- 4, and was slightly higher for < 12 month olds (85% +/- 5) versus children aged 1-5 years (74% +/- 5). 'Sedro tale nazel', which describes the chest as moving up and down, was a sensitive (86% +/- 3) and specific (60% +/- 4) indicator of chest indrawing. 'Tazyeek' (wheeze) had a sensitivity of 75% +/- 3 and specificity of 66% +/- 4 when compared to paediatrician assessment of wheezing during physical examination. Although not specific, the caretaker terms, 'nahagan' or 'nafas seria' (fast breathing) and 'sedro tale nazel' (chest indrawing), either spontaneously or after asking, were sensitive (> 71%) indicators of radiologic pneumonia and oxygen desaturation, and therefore can be used to prompt timely health seeking behaviour in these settings.

An instrument to assess acute respiratory infection case management in Egypt.

UNLABELLED: To develop an instrument to measure the quality of acute respiratory infection (ARI) case management among Egyptian children. METHODS: A baseline survey of all health facilities in a single district, using a multi-data source instrument. Data sources included providers, caretakers, patient records and observation of patient care. MAIN RESULTS: Physicians did not count the respiratory rate and check for subcostal retraction. Eighty-seven per cent of children who did not require antibiotics received them. Of five children who required antibiotics, four (80%) were prescribed an oral regimen. Three of these should have been admitted to a hospital but were not. Antibiotics were available at the facilities an estimated 7.9 months per year. Oxygen for inpatient treatment was available in one of two hospitals. CONCLUSIONS: This instrument was useful for comprehensively evaluating facility capability to provide quality case management. Deficiencies were identified but were not unexpected in a baseline survey. The Egypt ARI program has the potential to have a substantial impact on how children with ARI are diagnosed and treated in health facilities.

PIP: An instrument consisting of 10 questionnaires was developed to assess the ability of facilities to implement acute respiratory infection (ARI) case management guidelines. Data sources included interviews with physicians, nurses and area pharmacists; observation of patient care; review of patient records and an inventory of supplies. All 21 outpatient and two inpatient health facilities as well as 20 of 28 pharmacies in the district were included in the study. Of 93 child assessments observed, physicians asked the age for only 38 (41%). No child was questioned on ability to drink or experience of seizures. In addition, no physician checked for stridor, wheeze or chest indrawing, or counted the respiratory rate. 81 of 93 (87%) children with ARI were prescribed antibiotics. Among the 88 children assessed as not requiring antibiotics, 77 (88%) received them. Five children (2 pneumonia, 2 severe pneumonia, and 1 very severe disease) were determined to require antibiotics; four were prescribed an oral regimen. Of the four children that both required antibiotics and received them, three should have been admitted to the hospital for parenteral antibiotics, but were not. A variety of prescribed antibiotics were used among the 77 children. 27 (35%) children received two or three antibiotics without specifying the dose, frequency or duration on the prescription. Only two physicians mentioned the antibiotic dosage schedule for home care to the mother. Physicians at each of the outpatient facilities estimated the mean availability of antibiotics at 7.9 +or- 3.9 months. Three facilities (14%) had ampicillin suspension, none had amoxicillin and two (9%) had cotrimoxazole for the treatment of outpatient pneumonia. 19 (90%) had aspirin. Oxygen was available for inpatient care for children with pneumonia in one of the two hospitals, no nebulizers were available for treatment of wheezing, and disposable syringes were available in only one hospital. Parenteral bronchodilators were available in both, oral in neither. This instrument was useful for comprehensively evaluating facility capability to provide quality case management in the Egyptian ARI program.

Assessment of respiratory rate and chest indrawing in children with ARI by primary care physicians in Egypt.

In a baseline study for training purposes, two indicators of acute respiratory infections (the respiratory rate (RR) and chest indrawing) were assessed by Ministry of Health physicians in Egypt using a WHO test videotape. Chest indrawing, as defined by the WHO Acute Respiratory Infections (ARI) programme, was not widely recognized by current health personnel. Viewing a WHO training videotape led to significantly more correct assessments of chest indrawing compared with a group that had not viewed this videotape. The accuracy of using a timer versus a watch, and a 30-second versus 60-second counting interval was also evaluated. Rates counted over 60 seconds were more accurate than 30-second counts although the difference between them was not clinically significant. Counting of rates using timers with audible cues was comparable to using watches with second hands. Careful training of primary health workers in the assessment of RR and chest indrawing is essential if these clinical findings are to be used as reliable indicators in pneumonia treatment algorithms.

PIP: The timely treatment of acute lower respiratory tract infections (ALRI) in children depends upon the correct assessment of clinical findings. In the context of launching a national ALRI control program in Egypt, the authors surveyed the knowledge, attitudes, and practices of physicians in measuring respiratory rates (RR); examined the effect of a World Health Organization training videotape upon the recognition of chest indrawing; and examined the effect of different time intervals and providing audible timers on the accuracy of RR assessment. 320 Ministry of Health physicians participated in the study; 45% from maternal-child health care units, 30% from urban health centers, 18% in rural health units, and 7% in the district hospital or in school health. It was found that chest indrawing was not widely recognized by current health personnel and that the training compared with groups of physicians which were not exposed to the intervention. It was also found that counting RR for 60 seconds was more accurate than 30-second counts, but the difference between the 2 approaches was not clinically significant. Rate counting with audible cue timers was comparable to using watches with second hands. It is clear that primary health workers need to be carefully trained in assessing RR and chest indrawing if these clinical findings are to be used as reliable indicators in pneumonia treatment algorithms.