• View in gallery

    Seasonal pattern of pneumonia mortality and all other causes among Nairobi slum children. Poly = fractional polynomial transformation fitting line.

  • 1

    Black RE, Morris SS, Bryce J, 2003. Where and why are 10 million children dying every year. Lancet 361 :2226–2234.

  • 2

    Bryce J, Boschi-Pinto C, Shibuya K, Black RE, and WHO Child Health Epidemiology Reference Group, 2005. WHO estimates of the causes of death in children. Lancet 365 :1147–1152.

    • Search Google Scholar
    • Export Citation
  • 3

    Rudan I, Boschi-Pinto C, Biloglav Z, Mulholland K, Campbell H, 2008. Epidemiology and etiology of childhood pneumonia. Bull World Health Organ 86 :408–416.

    • Search Google Scholar
    • Export Citation
  • 4

    Kyobutungi C, Ziraba AK, Ezeh A, Ye Y, 2008. The burden of disease profile of residents of Nairobi’s slums: results from a demographic surveillance. Popul Health Metr 6 :1.

    • Search Google Scholar
    • Export Citation
  • 5

    Greenwood B, 2008. A global action plan for the prevention and control of pneumonia. Bull World Health Organ 86 :321–416.

  • 6

    APHRC, 2002 Health and Livelihood Needs of Residents of Informal Settlements in Nairobi City. Occasional Study Report No.1. Nairobi, Kenya.

  • 7

    Amuyunzu-Nyamongo M, Taffa N, 2004. The triad of poverty, environment and child health in Nairobi’s informal settlements. J Health Popul Dev Ctries 6 :1–14.

    • Search Google Scholar
    • Export Citation
  • 8

    Magadi MA, Zulu EM, Brockerhoff M, 2003. The inequality of maternal health care in urban sub-Saharan Africa in the 1990s. Population Studies 57 :347–366.

    • Search Google Scholar
    • Export Citation
  • 9

    Taffa N, 2003. A comparison of pregnancy and child health outcomes between teenage and adult mothers in the slums of Nairobi, Kenya. Int J Adolesc Med Health 15 :321–329.

    • Search Google Scholar
    • Export Citation
  • 10

    CBS, Ministry of Health, ORC Macro, 2004. Kenya Demographic and Health Survey (DHS) 2003. Calverton, MD: CBS, Ministry of Health, ORC Macro.

  • 11

    Yaka P, Sultan B, Broutin H, Janicot S, Philippon S, Fourquent N, 2008. Relationships between climate and year-to-year variability in meningitis outbreaks: a case study in Burkina Faso and Niger. Int J Health Geogr 7 :34.

    • Search Google Scholar
    • Export Citation
  • 12

    Sultan B, Labadi K, Guegan J, Janicot S, 2005. Climate drives the meningitis epidemics in West Africa. PLoS Med 2 :43–49.

  • 13

    Dowell S, Whitney C, Wright C, Schuchat A, 2003. Seasonal patterns of invasive pneoumococcal disease. Emerg Infect Dis 5 :573–578.

  • 14

    Kim P, Musher D, Glezen W, Rodriguez-Barradas M, Nahm W, Wright C, 1996. Association of invasive pneumococcal disease with season, atmospheric conditions, air pollution, and the isolation of respiratory viruses. Clin Infect Dis 22 :100–106.

    • Search Google Scholar
    • Export Citation
  • 15

    Roit I, 1994. Essential Immunology. Volume 8. Oxford: Blackwell Scientific Publications, Oxford, 331–333.

  • 16

    Salman H, Bergman M, Bessler H, Alexandrova S, Beilin B, Djaldetti M, 2000. Hypothermia affects the phagocytic activity of rat peritoneal macrophages. Acta Physiol Scand 168 :431–436.

    • Search Google Scholar
    • Export Citation
  • 17

    Van Loghem JJ, 1928. An epidemiological contribution to the knowledge of the respiratory system. J Hyg (Lond) 28 :33–54.

  • 18

    Tornheim J, Manya A, Oyando N, Kabaka S, Breiman R, Feikin D, 2007. The epidemiology of hospitalized pneumonia in rural Kenya: the potential of surveillance data in setting public health priorities. Int J Infect Dis 11 :536–543.

    • Search Google Scholar
    • Export Citation
  • 19

    Ndugwa RP, Zulu EM, 2008. Child morbidity and care-seeking in Nairobi slum settlements: the role of environmental and socioeconomic factors. J Child Health Care 12 :314–328.

    • Search Google Scholar
    • Export Citation
  • 20

    Ye Y, Louis V, Simboro S, Sauerborn R, 2007. Effect of meteorological factors on clinical malaria risk among children: an assessment using village-based meteorological stations and community-based parasitological survey. BMC Public Health 7 :101.

    • Search Google Scholar
    • Export Citation
  • 21

    Royston P, Ambler G, Sauerbrei W, 1999. The use of fractional polynomials to model continuous risk variables in epidemiology. Int J Epidemiol 28 :964–974.

    • Search Google Scholar
    • Export Citation
  • 22

    Royston P, 2000. A strategy for modeling the effect of a continuous covariate in medicine and epidemiology. Stat Med 19 :1831–1847.

  • 23

    Fullerton DG, Bruce N, Gordon SB, 2008. Indoor air pollution from biomass fuel smoke is a major health concern in the developing world. Trans R Soc Trop Med Hyg 102 :843–851.

    • Search Google Scholar
    • Export Citation
  • 24

    Smith KR, Samet JM, Romieu I, Bruce N, 2000. Indoor air pollution in developing countries and acute lower respiratory infections in children. Thorax 55 :518–532.

    • Search Google Scholar
    • Export Citation
  • 25

    Ezzati M, Kammen D, 2001. Indoor air pollution from biomass combustion and acute respiratory infections in Kenya: an exposure–response study. Lancet 358 :619–624.

    • Search Google Scholar
    • Export Citation
  • 26

    Awasthi S, Glick H, Fletcher R, 1996. Effect of cooking fuels on respiratory diseases in preschool children in Lucknow, India. Am J Trop Med Hyg 55 :48–51.

    • Search Google Scholar
    • Export Citation
  • 27

    Emuller E, Diab R, Binedell M, Hounsome R, 2003. Health risks of kerosene usage in an informal settlement in Durban, South Africa. Atmos Environ 36 :2015–2022.

    • Search Google Scholar
    • Export Citation
  • 28

    Tsai H, Kuo P, Liu C, Wang J, 2001. Respiratory viral infections among paediatric inpatients and outpatients in Taiwan from 1997 to 1999. J Clin Microbiol 39 :111–118.

    • Search Google Scholar
    • Export Citation
  • 29

    Kyobutungi C, Grau A, Stieglbauer G, Becher H, 2005. Absolute temperature, temperature changes and stroke risk: a case-crossover study. Eur J Epidemiol 20 :693–698.

    • Search Google Scholar
    • Export Citation
  • 30

    Quigley MA, 2005. Commentary: verbal autopsies–from small scale studies to mortality surveillance systems. Int J Epidemiol 34 :1087–1088.

    • Search Google Scholar
    • Export Citation
  • 31

    Etard JF, Le Hesran JY, Diallo A, Diallo JP, Ndiaye JL, Delaunay V, 2004. Childhood mortality and probable causes of death using verbal autopsy in Niakhar, Senegal, 1989–2000. Int J Epidemiol 33 :1286–1292.

    • Search Google Scholar
    • Export Citation
  • 32

    Korenromp EL, Williams BG, Gouws E, Dye C, Snow RW, 2003. Measurement of trends in childhood malaria mortality in Africa: an assessment of progress toward targets based on verbal autopsy. Lancet Infect Dis 3 :349–358.

    • Search Google Scholar
    • Export Citation
  • 33

    Snow RW, Armstrong JR, Forster D, Winstanley MT, Marsh VM, Newton CR, Waruiru C, Mwangi I, Winstanley PA, Marsh K, 1992. Childhood deaths in Africa: uses and limitations of verbal autopsies. Lancet 340 :351–355.

    • Search Google Scholar
    • Export Citation
  • 34

    Todd JE, de Francisco A, O’Dempsey TJ, Greenwood BM, 1994. The limitations of verbal autopsy in a malaria-endemic region. Ann Trop Paediatr 14 :31–36.

    • Search Google Scholar
    • Export Citation
  • 35

    Sazawal S, Black RE, and Pneumonia Case Management Trials Group, 2003. Effect of pneumonia case management on mortality in neonates, infants, and pre-school children: a meta-analysis of community based trials. Lancet Infect Dis 3 :547–556.

    • Search Google Scholar
    • Export Citation
  • 36

    Marsh DR, Gilroy KE, Van de Weerdt R, Wansi E, Qazi S, 2008. Community case management of pneumonia: at a tipping point? Bull World Health Organ 86 :381–389.

    • Search Google Scholar
    • Export Citation
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 

 

 

Seasonal Pattern of Pneumonia Mortality among Under-Five Children in Nairobi’s Informal Settlements

View More View Less
  • 1 African Population and Health Research Center (APHRC), Nairobi, Kenya

Using longitudinal data from the Nairobi Urban and Demographic Surveillance System (NUHDSS), we examined the seasonal pattern of pneumonia mortality among under-five children living in Nairobi’s slums. We included 17,787 under-five children resident in the NUHDSS from January 1, 2003 to December 31, 2005 in the analysis. Four hundred thirty-six deaths were observed and cause of death was ascertained by verbal autopsy for 377 of these deaths. Using Poisson regression, we modeled the quarterly mortality risk for pneumonia. The overall person-years (PYs) were 21,804 giving a mortality rate of 20.1 per 1,000 PYs in the study population. Pneumonia was the leading cause of death contributing 25.7% of the total deaths. Pneumonia mortality was highest in the second quarter (risk ratio [RR] = 2.3, confidence interval [CI]: 1.2–4.2 compared with the fourth quarter). The study provides evidence that pneumonia-related mortality among under-fives in Nairobi’s slums is higher from April to June corresponding to the rainy season and the beginning of the cold season.

INTRODUCTION

Acute lower respiratory tract infections (ALRIs—pneumonia, bronchiolitis, and bronchitis) are among the major killers of children less than 5 years of age in sub-Saharan Africa (SSA). Most of the deaths are caused by pneumonia. In SSA, pneumonia accounts for about 21% of the under-five mortality and therefore is the second killer after malaria, which accounts for 40% of under-five mortality.13 A recent study from Nairobi showed that pneumonia accounts for 3,463 years of life lost (YLLs) among under-fives in two Nairobi slums and accounts for 23% of the total YLLs among this age group.4

Despite the high burden of pneumonia, control of the disease, especially among young children, has had limited success and been relatively neglected.5 A number of reasons have been suggested for the neglect of the disease. For instance, most intervention programs do not give specific attention to the most vulnerable groups, such as deprived communities. The multiple aetiologies of pneumonia and lack of agreement among experts on the best intervention strategies have also undermined efforts to deal with the disease.

However, there are recent initiatives to put pneumonia control at the forefront of public health efforts. The Global Action Plan for Prevention and Control of Pneumonia (GAPP) is one such initiative. For this initiative to be successful there is a need to generate more research evidence on the epidemiology of the disease, especially among disadvantaged population such as slum dwellers, where pneumonia contributes substantially to the high disease burden and where underfives have a disproportionately high mortality burden, partly because of pneumonia.4 In slums, living conditions are characterized by poor housing (e.g., with walls mostly made of corrugated iron sheets), poor environmental hygiene, overcrowding, and poor access to health care. 6,7 These conditions impact dramatically on the health of slum dwellers. Compared with the rest of Kenya, residents of slum settlements exhibit worse health indicators, especially for the under-five population. 6,810

Seasonal variation of infectious diseases has been studied extensively in other settings. For instance, the incidence of meningococcal meningitis has been associated with the dry season in the Sahel region. 11,12 Additionally, several studies have investigated the effect of weather patterns—specifically cold weather–on the incidence of upper respiratory tract and oral tract infections, mostly in temperate countries. 13,14 Several pathophysiologic mechanisms have been suggested to explain how exposure to a sudden drop in temperature results in an acute respiratory infection. These include a reduction in the local immune response characterized by reduced phagocyte activity. The action of phagocytes in destroying viruses and bacteria is a major component of the non-specific immune response that is vital to prevent infection. 15,16 The failure of phagocytosis is exacerbated by the activation of latent sub-clinical infections in the population; hence, the sudden increase in incidence of acute respiratory infections that usually follows a sudden drop in temperature. 17 Apart from the purely biologic effect of cold weather, human response to weather changes may increase the risk of infection even further. Spending more time indoors in structures with poor ventilation and indoor pollution may further increase the risk of acute respiratory infections.

Seasonal variation of pneumonia has been studied in some African settings, mainly using hospital-based data. 18 However, there is little evidence of seasonal effects on pneumonia mortality at the population level. Slum dwellers, despite their high morbidity and mortality burden, have low health care utilization rates 6,7,19; hence, hospital-based data may not capture the true burden of the disease or its seasonality effects, if any, among this population. Given the poor housing and environmental conditions in the slums settlements, one may not necessarily expect to observe exceptionally high seasonal variation in pneumonia-related mortality because children in these communities perpetually live in conditions that may predispose them to the risk of contracting ALRIs.

Against this background, this work uses longitudinal data from the Nairobi Urban and Demographic Surveillance System (NUHDSS), to determine whether there is a seasonal pattern of mortality caused by pneumonia among children less than 5 years of age living in two of Nairobi’s informal settlements.

METHODS

Study site.

The study was conducted in two informal settlements of Korogocho and Viwandani in the city of Nairobi, where the African Population and Health Research Center (APHRC) is running the NUHDSS. The settlements are located about 5 to 10 km from the city center and 3 km from each other. The population under surveillance on January 1, 2007 was 59,513 individuals living in 21,993 households. Poor environmental sanitation, overcrowded houses, and poor access to basic health care characterize these settlements. 6,7

Nairobi is situated at a high altitude of 1,700 meters above sea level. This geographic situation makes Nairobi relatively cold compared with other tropical cities. The average annual minimum temperature is 12°C (range: 11–14°C) and average maximum temperature is 23°C (range: 21–26°C). There are two rainy seasons in a year in Nairobi. The first rainy season is from March to May and the second from October to December. The total annual rainfall is about 900 mm, with the most rainfall in the second quarter of the year (April = 199 mm, May = 155 mm). The second and third quarters of the year correspond to the cold season, but temperatures are much lower in the third quarter (mean temperature 17°C with a minimum temperature of 11°C).

Study subjects.

The study involved 17,787 children less than 5 years of age who lived at one time or another in the NUHDSS area from January 1, 2003 to December 31, 2005. Children entered the NUHDSS either by enumeration at the initial census or through birth or immigration. The 17,787 children contributed 21,804 person years (PYs) of exposure that were used in the analysis.

Ascertaining cause of death.

Data on key demographic events (birth, death, migratory movements) are collected every 4 months by trained and experienced interviewers among all NUHDSS residents. For all recorded deaths, the cause of death is ascertained using the verbal autopsy (VA) approach. Details of how the method is applied in the NUHDSS have been described elsewhere.4 Briefly, a VA questionnaire is administered to a credible respondent, preferably a parent who took care of the sick child before death. Completed VA forms are reviewed independently by three physicians to determine the most probable cause of death. If the assigned code is similar for at least two of the physicians, the cause of death is assigned. If not, the most probable cause of death is assigned at consensus meetings among the three physicians. Where disagreements persist, the cause of death is coded as undetermined. Causes of death are classified according to the International Classification of Disease Version 10 (ICD10), using a modified and abridged code list that can be mapped onto the global burden of disease cause list. On this modified code list, uncommon causes of death in the study area are collapsed in broader categories such as “other specified communicable diseases” to cater for rare communicable diseases such as typhoid fever.

Analysis.

The time spent under observation by the study population was calculated for each month of the year starting from January 1, 2003. The number of days an individual was resident in the NUHDSS during a particular month was obtained from the NUHDSS database. These days were then summed up for all individuals and person months calculated. To obtain the total person months for the 3 years in the study period, we summed up the individual contribution for the same months in all the years.

Because of the small number of deaths, monthly mortality rates were calculated by summing up the number of deaths during the same month over the 3 years and dividing it by the corresponding person time. We calculated the monthly mortality rates using Poisson regression with the logarithm of person time as offset. This was done separately for pneumonia, for all other causes combined and for non-infectious causes excluding birth-related deaths. To assess the relationship between mortality rate and season, we used the fractional polynomial approach through which the best fitting transformation of the linear dependent variable is determined. 2022 The best fit was reached with a first degree quadratic transformation for pneumonia, all other causes and non-infectious causes.

To compute the seasonal mortality risk for pneumonia and for all other causes of death, we fit a Poisson regression model controlling for calendar year (2003–2005), site of residence, and age (below 1 year and 1 to 4 years). Risk ratios (RR) were computed for each quarter of the year in comparison to the fourth quarter (October–December), which had the lowest number of pneumonia deaths.

RESULTS

Study population characteristics.

In total, 17,787 children less than 5 years of age resident in the NUHDSS from January 1, 2003 to December 31, 2005 participated in the study. The children contributed 21,805 PYs over the 3 years. There was slight variation in the distribution of PYs between quarters. The highest contribution was for the fourth quarter (6,017 PYs) and the lowest for the first quarter (4,826 PYs). Among other covariates, significant differences were observed in the age distribution of PYs whereby a larger proportion of observation time was accumulated in the age group above 1 year (16,457 PYs or 75.5% of the overall PYs). The distribution of person time between categories of sex, site, and calendar year is shown in Table 1.

Cause of death.

Among 436 deaths observed, cause of death was assigned in 377 cases (86.5%) and missing in 59 cases (13.5%), mostly caused by inability to administer the VA (71.2%) and incomplete coding (28.8%) cases. Pneumonia was the leading cause of death, contributing 25.7% of deaths, whereas diarrhea contributed 22.0%. The contribution of other individual causes was small and ranged from 0.3% to 6.9% (Table 2).

Monthly mortality from pneumonia and all other causes.

Pneumonia-related mortality followed the same seasonal pattern as the mortality from all other causes. The mortality was highest in the second quarter of the year especially in June (60.1 per 100,000 PYs). From the third quarter, there was a sharp decrease in the mortality from all other causes. Pneumonia-related mortality also decreased in the third quarter, but with a less steep gradient. In fact, mortality remained high in the first 2 months of the third quarter (July: 49.4 per 100,000 PYs; August: 39.3 per 100,000 PYs) and declined only in September. In the last quarter of the year, the pneumonia mortality rate reached its lowest level and the lowest rate was observed in November (16.4 per 100,000 PYs). Similarly, mortality from all other causes was lowest in the fourth quarter of the year (Table 3 and Figure 1).

Seasonal effect on pneumonia mortality.

There was a significant association between season (expressed as quarter) and pneumonia mortality in the study area. Compared with the fourth quarter of the year and after controlling for calendar year, site, and age, there was significantly higher mortality from pneumonia in the second and the third quarters of the year (Quarter 2: RR = 2.1, 95% CI: 1.1–3.9; Quarter 3: RR = 1.9, 95% CI: 1.0–3.6). The risk in the first quarter of the year was also higher than in the fourth but not significantly so, RR = 1.8, 95% CI: 0.9–3.4. Mortality from all other causes was also significantly higher in the second and third quarters compared with the fourth quarter (Table 4). In contrast, there was no significant difference between mortality from non-infectious conditions (excluding birth-related deaths) in the first, second, and third quarters compared with the fourth quarter (RR 1.1, 1.2, and 1.1, respectively). A peak in mortality from birth-related deaths was observed in July; however, this also coincided with a peak in the number of births (results not shown).

Among the control variables, site of residence and age were significantly associated with pneumonia mortality. The risk of pneumonia mortality was higher in Korogocho compared with Viwandani (RR = 2.2, 95% CI: 1.4–3.3). As would be expected, children between 1 and 4 years of age experienced less risk of dying from pneumonia compared with children less than 1 year (RR = 0.1, 95% CI: 0.1–0.2) (Table 4).

DISCUSSION

Using longitudinal and verbal autopsy data from the NUHDSS, this study showed a high burden of pneumonia as reported in another study conducted in the same area.4 It also showed that pneumonia-related mortality rates are not constant throughout the year. A seasonal pattern is observed whereby mortality in the third quarter of the year is significantly higher than in the fourth quarter.

A possible explanation for the high pneumonia mortality is the poor housing and overcrowding (about 60,000 individuals live on < 1 km2), conditions that facilitate transmission of the pneumonia-causing agents. In addition, indoor air pollution could be another important contributing factor, because it has been reported widely as a major risk factor for acute upper and lower respiratory tract infections, especially among under-fives. 2325 In the slums, most residents use the same room for cooking and sleeping and 95% of households use kerosene stoves for cooking. The particulate matter and carbon dioxide emitted by burning kerosene are inflammatory agents that lead to breaches in the epithelial lining of the respiratory tract, hence increasing the risk of respiratory tract infections. 26,27 There is a need to investigate further the role of indoor air pollution on the risk of acute respiratory tract infections among the slum population, particularly among under-five children, and the impact of improving indoor air quality on childhood mortality. Such an investigation could involve indoor monitoring of air quality and levels of particulate matter.

The seasonal pneumonia mortality pattern corresponds to the temperature pattern throughout the year, with higher mortality observed during the cold months and lower mortality during the warm months. However, the highest monthly mortality was observed in June, whereas the coldest month was August. What is peculiar about June is the fact that it marks the beginning of the cold season, which lasts until August. Studies have suggested that the risk for respiratory tract infections is relatively high soon after a sudden drop in temperature, as may happen in the transition from a warm/hot season to a cold one. Seasonality of pneumonia, other ARTIs, and other viral infections have been observed in many temperate countries, especially in association with winter. 13,28 It has been suggested that such seasonality does not exist in tropical countries because changes in temperature between hot and cold seasons are not as dramatic as in temperate countries. However, since Nairobi is a city at a high altitude, it is different from most other tropical settings. Indeed, monthly temperature and rainfall for the years 2003–2005 for the city of Nairobi (http://iri.columbia.edu/climate/forecast) show that, although the mean annual temperature is moderate (24°C), there is a sudden drop of temperature in the second and third quarters of the year, with minimum temperature ranging between 11 and 14°C. Such conditions could mimic temperate climates.

Because overall mortality showed a similar seasonal pattern, we examined the mortality from other non-infectious health conditions (56 cases from 12 causes) and found no such seasonal pattern. The RR for non-infectious conditions excluding birth-related deaths in the first, second, and third quarters compared with the fourth quarter were all close to 1.0 and non-significant suggesting that the seasonality observed for pneumonia does not occur for all the other causes of death.

Korogocho children have significantly higher pneumonia mortality than those in Viwandani. Viwandani is mostly inhabited by labor migrants seeking employment in the nearby industrial area. Although there are more employment opportunities there, it is also a cash-based community. As a result, a larger proportion of residents in Viwandani stay for short periods and move on compared with Korogocho. It is possible that either residents do not stay long enough to be exposed to the hazardous slum environment or that, economically unsuccessful migrants, who could potentially have worse health outcomes move elsewhere and leave behind more successful ones resulting in selection of the stable Viwandani population. Detailed studies to explain this discrepancy are underway.

Limitations.

Our study has some limitations. By linking season with pneumonia mortality, the study does not take into account actual individual exposure. Some individuals may not have been physically resident in the study area all year, especially during the high-risk periods or those who were resident could have reacted differently to the weather/weather changes and hence altered their risk of exposure.

The yearly quarters represent changes in seasons; however, we did not link actual changes in temperature to pneumonia mortality. In addition, since a quarter is 3 months long, there may be wide variation in the daily temperature over that period. Such variations are masked in our analysis and may explain the moderate associations with seasons. Further research should aim to make this link more explicit, for instance by using a case-crossover design with daily temperature measures as the main explanatory variable. 29

Several questions have been raised about the accuracy of the VA methods to classify causes of death correctly, especially for under-fives. 3033 In this study, we acknowledge the possibility of misclassifications of some causes of death. On the other hand, in an area of high human immunodeficiency virus (HIV) burden, it is likely that some acquired immunodeficiency syndrome (AIDS)-related deaths were classified as pneumonia, especially if there was co-infection. In a setting where most deaths occur outside the formal health care system, where care is not always sought before death, the verbal autopsy tool is currently the best alternative of assigning cause of death despite its shortcomings. Unless the HIV serostatus had been established before death or other social history was available in the VA interview, it is not possible to attribute death to HIV-related conditions in this population. This misclassification may have led to the overestimation of the overall pneumonia mortality. However, our estimates of seasonal risk are likely to be robust, because we do not expect the specificity of assigning pneumonia as the cause of death to vary across seasons, as is the case for malaria, 34 because pneumonia has very distinct signs and symptoms.

We assessed the seasonal pattern based on 3-year data; one may argue that the pattern may change if we have data from more years, given that there may be annual variation in weather patterns. However, given that in our model the variable “year” did not have any effect on the risk of pneumonia mortality, we can say that the estimates obtained are likely to be the same with data for more years.

CONCLUSION

The results of this study are of public health interest and should inform some changes in policy and practice in Nairobi City. Although we advocate more research to validate our study findings, we believe that there is a need for action to address pneumonia mortality. Pneumonia deaths can be prevented in several ways, including early detection and case management, and vaccination. Implementation of existing guidelines in the integrated management of childhood illnesses and introduction of pneumococcal and Hib vaccines could go a long way in reducing childhood mortality among slum dwellers in Nairobi. Special interventions could also be put in place just before or soon after the onset of the cold seasons not only to cater for pneumonia but for other causes of childhood mortality. Outreach programs and case-finding activities by community workers during cold spells may increase the early case-detection rates and treatment of common childhood conditions.

Although the suggested preventive strategies are likely to be effective, they will not be able to prevent all the pneumonia cases. There is a need to put in place effective case management systems that include early detection and treatment of cases with antibiotics. Unfortunately, this approach may be challenging in a slum setting where access to formal health care still is limited. An alternative would be to provide treatment within the community, an approach that has been proven to reduce mortality and morbidity from pneumonia, 35 throughout the year but particularly during the high-risk season, despite concerns about lack of trained staff to administer the antibiotics. 36

Table 1

Characteristics of study participants

Table 1
Table 2

Causes of death among under-five children in Korogocho and Viwandani, 2003–2005*

Table 2
Table 3

Monthly mortality rates from pneumonia and all other causes (including unknown)*

Table 3
Table 4

Pneumonia and all other causes—including unknown causes*

Table 4
Figure 1.
Figure 1.

Seasonal pattern of pneumonia mortality and all other causes among Nairobi slum children. Poly = fractional polynomial transformation fitting line.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 81, 5; 10.4269/ajtmh.2009.09-0070

*

Address correspondence to Yazoume Ye, African Population and Health Research Center (APHRC), Shelter Afrique Center, 2nd Flr, Longonot Road, Upper Hill, PO Box 10787 00100, GPO, Nairobi, Kenya. E-mails: yyazoume@aphrc.org or yyazoume@hotmail.com

Authors’ addresses: Yazoume Ye, Eliya Zulu, Maurice Mutisya, Benedict Orindi, Jacques Emina, and Catherine Kyobutungi, African Population and Health Research Center (APHRC), Shelter Afrique Center, 2nd Flr, Longonot Road, Upper Hill, PO Box 10787 00100, GPO, Nairobi, Kenya, Tel: +254-20-2720400/1/2, Fax: +254-20-2720380, E-mails: yyazoume@aphrc.org, ezulu@aphrc.org, mmutisya@aphrc.org, borindi@gmail.com, jemina@aphrc.org, and ckyobutungi@aphrc.org.

Acknowledgments: We acknowledge the contribution of APHRC’s dedicated field and data management teams. We also acknowledge the contribution of Alex Ezeh to the conceptualization of the NUHDSS.

Financial support: The authors are supported by a grant from the Wellcome Trust UK (grant number GR078530AIA). Work in the NUHDSS has been supported by grants from the Rockefeller Foundation and the William and Flora Hewlett Foundation.

REFERENCES

  • 1

    Black RE, Morris SS, Bryce J, 2003. Where and why are 10 million children dying every year. Lancet 361 :2226–2234.

  • 2

    Bryce J, Boschi-Pinto C, Shibuya K, Black RE, and WHO Child Health Epidemiology Reference Group, 2005. WHO estimates of the causes of death in children. Lancet 365 :1147–1152.

    • Search Google Scholar
    • Export Citation
  • 3

    Rudan I, Boschi-Pinto C, Biloglav Z, Mulholland K, Campbell H, 2008. Epidemiology and etiology of childhood pneumonia. Bull World Health Organ 86 :408–416.

    • Search Google Scholar
    • Export Citation
  • 4

    Kyobutungi C, Ziraba AK, Ezeh A, Ye Y, 2008. The burden of disease profile of residents of Nairobi’s slums: results from a demographic surveillance. Popul Health Metr 6 :1.

    • Search Google Scholar
    • Export Citation
  • 5

    Greenwood B, 2008. A global action plan for the prevention and control of pneumonia. Bull World Health Organ 86 :321–416.

  • 6

    APHRC, 2002 Health and Livelihood Needs of Residents of Informal Settlements in Nairobi City. Occasional Study Report No.1. Nairobi, Kenya.

  • 7

    Amuyunzu-Nyamongo M, Taffa N, 2004. The triad of poverty, environment and child health in Nairobi’s informal settlements. J Health Popul Dev Ctries 6 :1–14.

    • Search Google Scholar
    • Export Citation
  • 8

    Magadi MA, Zulu EM, Brockerhoff M, 2003. The inequality of maternal health care in urban sub-Saharan Africa in the 1990s. Population Studies 57 :347–366.

    • Search Google Scholar
    • Export Citation
  • 9

    Taffa N, 2003. A comparison of pregnancy and child health outcomes between teenage and adult mothers in the slums of Nairobi, Kenya. Int J Adolesc Med Health 15 :321–329.

    • Search Google Scholar
    • Export Citation
  • 10

    CBS, Ministry of Health, ORC Macro, 2004. Kenya Demographic and Health Survey (DHS) 2003. Calverton, MD: CBS, Ministry of Health, ORC Macro.

  • 11

    Yaka P, Sultan B, Broutin H, Janicot S, Philippon S, Fourquent N, 2008. Relationships between climate and year-to-year variability in meningitis outbreaks: a case study in Burkina Faso and Niger. Int J Health Geogr 7 :34.

    • Search Google Scholar
    • Export Citation
  • 12

    Sultan B, Labadi K, Guegan J, Janicot S, 2005. Climate drives the meningitis epidemics in West Africa. PLoS Med 2 :43–49.

  • 13

    Dowell S, Whitney C, Wright C, Schuchat A, 2003. Seasonal patterns of invasive pneoumococcal disease. Emerg Infect Dis 5 :573–578.

  • 14

    Kim P, Musher D, Glezen W, Rodriguez-Barradas M, Nahm W, Wright C, 1996. Association of invasive pneumococcal disease with season, atmospheric conditions, air pollution, and the isolation of respiratory viruses. Clin Infect Dis 22 :100–106.

    • Search Google Scholar
    • Export Citation
  • 15

    Roit I, 1994. Essential Immunology. Volume 8. Oxford: Blackwell Scientific Publications, Oxford, 331–333.

  • 16

    Salman H, Bergman M, Bessler H, Alexandrova S, Beilin B, Djaldetti M, 2000. Hypothermia affects the phagocytic activity of rat peritoneal macrophages. Acta Physiol Scand 168 :431–436.

    • Search Google Scholar
    • Export Citation
  • 17

    Van Loghem JJ, 1928. An epidemiological contribution to the knowledge of the respiratory system. J Hyg (Lond) 28 :33–54.

  • 18

    Tornheim J, Manya A, Oyando N, Kabaka S, Breiman R, Feikin D, 2007. The epidemiology of hospitalized pneumonia in rural Kenya: the potential of surveillance data in setting public health priorities. Int J Infect Dis 11 :536–543.

    • Search Google Scholar
    • Export Citation
  • 19

    Ndugwa RP, Zulu EM, 2008. Child morbidity and care-seeking in Nairobi slum settlements: the role of environmental and socioeconomic factors. J Child Health Care 12 :314–328.

    • Search Google Scholar
    • Export Citation
  • 20

    Ye Y, Louis V, Simboro S, Sauerborn R, 2007. Effect of meteorological factors on clinical malaria risk among children: an assessment using village-based meteorological stations and community-based parasitological survey. BMC Public Health 7 :101.

    • Search Google Scholar
    • Export Citation
  • 21

    Royston P, Ambler G, Sauerbrei W, 1999. The use of fractional polynomials to model continuous risk variables in epidemiology. Int J Epidemiol 28 :964–974.

    • Search Google Scholar
    • Export Citation
  • 22

    Royston P, 2000. A strategy for modeling the effect of a continuous covariate in medicine and epidemiology. Stat Med 19 :1831–1847.

  • 23

    Fullerton DG, Bruce N, Gordon SB, 2008. Indoor air pollution from biomass fuel smoke is a major health concern in the developing world. Trans R Soc Trop Med Hyg 102 :843–851.

    • Search Google Scholar
    • Export Citation
  • 24

    Smith KR, Samet JM, Romieu I, Bruce N, 2000. Indoor air pollution in developing countries and acute lower respiratory infections in children. Thorax 55 :518–532.

    • Search Google Scholar
    • Export Citation
  • 25

    Ezzati M, Kammen D, 2001. Indoor air pollution from biomass combustion and acute respiratory infections in Kenya: an exposure–response study. Lancet 358 :619–624.

    • Search Google Scholar
    • Export Citation
  • 26

    Awasthi S, Glick H, Fletcher R, 1996. Effect of cooking fuels on respiratory diseases in preschool children in Lucknow, India. Am J Trop Med Hyg 55 :48–51.

    • Search Google Scholar
    • Export Citation
  • 27

    Emuller E, Diab R, Binedell M, Hounsome R, 2003. Health risks of kerosene usage in an informal settlement in Durban, South Africa. Atmos Environ 36 :2015–2022.

    • Search Google Scholar
    • Export Citation
  • 28

    Tsai H, Kuo P, Liu C, Wang J, 2001. Respiratory viral infections among paediatric inpatients and outpatients in Taiwan from 1997 to 1999. J Clin Microbiol 39 :111–118.

    • Search Google Scholar
    • Export Citation
  • 29

    Kyobutungi C, Grau A, Stieglbauer G, Becher H, 2005. Absolute temperature, temperature changes and stroke risk: a case-crossover study. Eur J Epidemiol 20 :693–698.

    • Search Google Scholar
    • Export Citation
  • 30

    Quigley MA, 2005. Commentary: verbal autopsies–from small scale studies to mortality surveillance systems. Int J Epidemiol 34 :1087–1088.

    • Search Google Scholar
    • Export Citation
  • 31

    Etard JF, Le Hesran JY, Diallo A, Diallo JP, Ndiaye JL, Delaunay V, 2004. Childhood mortality and probable causes of death using verbal autopsy in Niakhar, Senegal, 1989–2000. Int J Epidemiol 33 :1286–1292.

    • Search Google Scholar
    • Export Citation
  • 32

    Korenromp EL, Williams BG, Gouws E, Dye C, Snow RW, 2003. Measurement of trends in childhood malaria mortality in Africa: an assessment of progress toward targets based on verbal autopsy. Lancet Infect Dis 3 :349–358.

    • Search Google Scholar
    • Export Citation
  • 33

    Snow RW, Armstrong JR, Forster D, Winstanley MT, Marsh VM, Newton CR, Waruiru C, Mwangi I, Winstanley PA, Marsh K, 1992. Childhood deaths in Africa: uses and limitations of verbal autopsies. Lancet 340 :351–355.

    • Search Google Scholar
    • Export Citation
  • 34

    Todd JE, de Francisco A, O’Dempsey TJ, Greenwood BM, 1994. The limitations of verbal autopsy in a malaria-endemic region. Ann Trop Paediatr 14 :31–36.

    • Search Google Scholar
    • Export Citation
  • 35

    Sazawal S, Black RE, and Pneumonia Case Management Trials Group, 2003. Effect of pneumonia case management on mortality in neonates, infants, and pre-school children: a meta-analysis of community based trials. Lancet Infect Dis 3 :547–556.

    • Search Google Scholar
    • Export Citation
  • 36

    Marsh DR, Gilroy KE, Van de Weerdt R, Wansi E, Qazi S, 2008. Community case management of pneumonia: at a tipping point? Bull World Health Organ 86 :381–389.

    • Search Google Scholar
    • Export Citation
Save