Institute of Child Health and Hospital for Children, Chennai is a referral pediatric institute and a 537 bedded multi-specialty hospital attached to Madras Medical College, Chennai. It caters to health needs of children up to 12 years of age from the city of Chennai and neighbouring districts. We conducted a retrospective chart review of all children in the age group of 1 month to 59 months admitted between January 2006 and December 2008 in our hospital with a final diagnosis of pneumonia. These cases were classified in to 3 groups (i) clinically diagnosed pneumonia {WHO diagnostic criteria for pneumonia (fast breathing defined by respiratory rate ≥60/min in <2 months of age, ≥50/min in 2- 11 months of age and ≥40/min in 12-59 months of age) or severe pneumonia (pneumonia with chest retraction) as per IMCI guidelines} [12], (ii) radiologically diagnosed pneumonia (based on the consensus finding of chest infiltrate(s) by a qualified radiologist and treating clinician), and (iii) clinically and radiologically diagnosed pneumonia (when both criteria were present). Children whose final diagnosis was tuberculous pneumonia or chemical pneumonitis following kerosene ingestion were excluded. Those in whom the data were incomplete (discharged against medical advice, absconded) were also excluded.

A pre-designed proforma was used to record the demographic profile (age and sex), symptoms (fever, cough, difficulty in breathing, altered sensorium, convulsions, inability to feed and wheezing), physical findings (weight, temperature, level of consciousness, cyanosis, shock, respiratory rate, chest retraction, grunting, auscultatory signs such as crackles and wheeze, associated congenital heart disease), radiologic findings(chest radiograph findings were available for all cases), oxygen saturation, need for assisted ventilation and final outcome. The possible outcomes were "Discharged"(those children who recovered) as per the treating physician or "Death". Risk factors for mortality analyzed were young age of 1-6 months, female sex, presence of wheeze, respiratory rate ≥70/min, chest retraction (lower chest indrawing), altered level of consciousness (ALOC) as denoted by unresponsiveness to voice or pain, convulsion at the time of presentation, shock as defined by Pediatric Advanced Life Support (PALS) guidelines [13], associated congenital heart disease (CHD) confirmed by echocardiogram, measles infection within past three months, weight for age <-2 Z score, and need for assisted ventilation at the time of presentation. Need for assisted ventilation was based on the treating clinician’s decision. Nutritional status was recorded by using the Z score as per WHO standards of weight for age [14]. Risk factors for all the three study groups of pneumonia were analyzed separately.

In the period between January 2006 and December 2008, there were 4976 children admitted in the age group of 1month to 59 months with a final diagnosis of pneumonia. 4375 children (42% girls) were included in the study after excluding those with chemical pneumonitis or tuberculous pneumonia and those in whom the data were incomplete There were 1536 children who were diagnosed to have pneumonia based on clinical features alone, 1020 children who were diagnosed by radiological findings alone and 1819 children who had clinical as well as radiological evidence of pneumonia (Fig. 1). Total cases of pneumonia constituted 8% of hospital admissions in this period. There were 1176, 1670 and 1529 children with pneumonia in the years 2006, 2007 and 2008, respectively.

Fig.1 Flow chart of study subjects.

Majority of the children (2175; 48%) were in the age group 1-6 months (Table I). 357 children died providing an over all case fatality rate of 8.2 % (95% CI: 7.37 - 8.99%). The all cause mortality in the age group of 1 month to 59 months was 3.9% in this period as per the hospital statistics.

TABLE I Age and Case Fatality Rate Among Pneumonia Cases

The difference in mortality due to pneumonia among male and female children was not statistically significant (P=0.12). The mortality rate was lower in older children as compared to younger ones (P=0.001) (Table I). The difference in CFR among the three study groups was not significant, except in clinically and radiologically diagnosed cases in the age group of 24 to 59 months, which was higher compared to clinically diagnosed and radiologically diagnosed cases (Table II).

TABLE II Comparison of Case Fatality Among Study Groups

On multivariate analysis, need for assisted ventilation was found to be the only independent risk factor for death among all the three study groups of pneumonia. In addition young age of 1-6 months, weight for age < - 2Z score, altered level of consciousness (ALOC) and CHD among clinically diagnosed pneumonia cases; ALOC and CHD among radiologically diagnosed pneumonia cases; and weight for age <-2Z score among clinically and radiologically diagnosed pneumonia cases were found to be significant independent risk factors for death (Tables III, IV and V).

TABLE III Risk Factors for Mortality Among Clinically Diagnosed Pneumonia Cases

TABLE IV Risk Factors for Mortality Among Radiologically Diagnosed Pneumonia Cases

TABLE V Risk Factors for Mortality Among Clinically and Radiologically Diagnosed Pneumonia Cases

There was a trend of increase in death rate with worsening nutritional status (Table VI). Poor nutritional status was strikingly associated with mortality in children younger than 24 months of age, as mortality was seen only in children with - Z score in this age group.

TABLE VI	Nutritional Status and Outcome in Pneumonia 

Averting pneumonia deaths will significantly contribute to realizing Millennium Development Goal 4 (MDG4) of reducing under five mortality [13]. This retrospective analysis was conducted to identify the demographic and clinical variables associated with deaths in hospitalized children of 1 month –59 months of age with a diagnosis of pneumonia. The case fatality rate (CFR) of pneumonia was 8.2% compared to 3.9% for all cause mortality in this age group in the same period. CFR of childhood pneumonia in various Indian studies ranges between 8.9% to 47% [4,5,9,10,15,16,] and 3.4% to 12% in other developing countries [6,7,17].

The risk factors for deaths due to pneumonia vary between countries, regions and communities. Frequently studied factors are young age, low birth weight, under nutrition, anemia, lack of parental education, over-crowding, indoor air pollution, lack of exclusive breast feeding, lack of measles immunization, co-morbidities such as congenital heart diseases and other congenital anomalies and severity at presentation to hospital [4-9,11]. In this study, need for assisted ventilation was found to be independently associated with fatal pneumonia of clinically, radiologically, as well as clinically and radiologically diagnosed pneumonia cases. Need for assisted ventilation at the time of presentation is likely to indicate the severity. In another study, cyanosis and head nodding were the determining factors for mechanical ventilation [4]. Other factors like young age of 1- 6 months, weight for age <-2 Z score, altered level of consciousness and CHD were found to be associated with mortality in pneumonia, even though these factors were not observed in all the three study groups. The different combination of risk factors among these groups may be due to ambiguity in clinical features for pneumonia, associated wheeze, CHD, etc. Young age has been reported to be a risk factor for mortality in studies from India and other developing countries [7-9, 11]. Malnutrition was another significant contributory factor for mortality in other studies also [9,11, 17]. Not only that malnutrition was associated with increased fatality, there was an incremental effect in pneumonia mortality with increasing degree of malnutrition when all the cases were considered together. CHD was a risk factor for pneumonia mortality in a study [7] while not so in another study [4].A more detailed analysis with reference to the type(cyanotic or acyanotic) and severity of CHD is needed to clarify its role in pneumonia deaths.

Limitations of the study are that specific etiology of pneumonia could not be identified by blood culture or other methods and radiological diagnosis was by the radiologist and clinician by consensus and not as per WHO criteria [18]. Other potentially relevant risk factors such as lack of exclusive breast feeding, indoor air pollution, hypoxia, increased C-reactive protein and head nodding were not studied due to non availability of these data in view of retrospective nature of the study. The sample studied also was less than calculated sample size. A strength of the study was the study period of three years preventing the effect of epidemic outbreak and seasonality.

Case fatality rate among children 1 month to 59 months of age hospitalized with community acquired pneumonia was 8.2%. Need for assisted ventilation was found to be independently associated with mortality in children aged 1-59 months admitted with pneumonia. Young age (1- 6 months), malnutrition (<-2 Z weight for age), altered level of consciousness and associated congenital heart disease are other factors and hence more aggressive monitoring and management are required in children with these risk factors.

Contributors: PR: Conceived, designed the study and collected the data. PR and KN: Drafted the article. KN: Revised it critically for important intellectual content. AV: managed data, analyzed and interpreted the results. SS: Co-ordinated and guided the research.

Funding: Nil; Competing interests: None stated.