Mortality within 24 hours of admission to the Paediatric Unit, Harare Central Hospital, Zimbabwe.

OBJECTIVE: To determine the proportion of deaths, characteristics of children and risk factors for mortality w ithin 24 hours of admission to a Paediatric hospital in Harare. STUDY DESIGN: Prospective cohort study. SETTINGS: Paediatric Unit, Harare Central Hospital. SUBJECTS: All patients admitted to the medical wards who consented to participate were enrolled. Preadmission factors including duration of illness and health seeking behaviour prior to presentation, delays in A&E department assessed by lag time to assessment, administration of initial medications and admission to the ward were documented. The presenting clinical signs and admission diagnoses were also recorded MAIN OUTCOME MEASURE: Death within 24 hours of admission. RESULTS: Of the 737 paediatric admissions during the study period, 54 children died within 24 hours giving a case fatality rate of 7.3%. These constituted 34.6% of total deaths in the study population (54/155). The median age of the children in this study was 16 months (Q1 = 4, Q3 = 36) and 53.2% were male. Having subcostal recessions on admission was significantly associated with mortality (within 24 hours of admission) with a RR 29.9 (95% CI 1.56.74) while socio-demographic factors, duration of illness, fever, diagnosis on admission and delays in A & E department were not. CONCLUSION: The contribution of deaths within 24 hours of admission to the overall mortality in children remains unacceptably high. Sub-costal recessions on admission (a proxy for severe pneumonia) had the highest risk of mortality within 24 hours of admission. There is need for early identification and aggressive management of children with pneumonia.

The levels of malnutrition and risk factors for mortality at Harare Central Hospital-Zimbabwe: an observation study.

OBJECTIVES: To determine the prevalence of malnutrition, and identify risk factors associated with mortality in acute severe malnutrition in a major referral hospital in Harare. DESIGN: Cross sectional analytical study. SETTING: Harare Central Hospital, paediatric wards. SUBJECTS: All children admitted to the general paediatric medical wards between 12 October 2003 and 19 January 2004 were surveyed. MAIN OUTCOME MEASURES: Prevalence and mortality in hospitalized children with acute severe malnutrition. STUDY FACTORS: Patient's age, sex, vaccination status, type of malnutrition, weight-for-height, breast feeding status (age <24 months), care giver details, orphanage, area of residence, new or re-admission, time of admission, admission temperature, co-morbidity conditions, HIV status and selected laboratory tests. RESULTS: A total of 784 infants and children were admitted during the study period, of whom 619 were eligible for the study. Of the 619 children, 259 (41.8%) had acute severe malnutrition, 79 (12.8%) moderate malnutrition and 281 (45.5%) had no malnutrition. Fatality rates were 42.9% (acute severe malnutrition), 32.9% (moderate malnutrition), and 21% (no malnutrition) respectively. Factors predictive of mortality by multivariate analysis were age <18 months (O.R=2.27; 95% CI 1.20-4.29), weight-for-height <70% (O.R=2.63; 95% CI=1.24-5.56), acute diarrhoea (O.R=3.42; 95% CI=1.53-7.65), persistent diarrhoea (O.R=2.67; 95% CI= 1.26-5.66), and pneumonia (O.R=2.21; 95% CI= 1.08-4.52). CONCLUSION: Mortality among children with acute severe malnutrition at this institution was unacceptably high. Case management needs strengthening particularly for malnutrition, diarrhoea and pneumonia. The role of high HIV prevalence rates on mortality in this population needs evaluating.

Long-term outcome of preterm infants discharged home on kangaroo care in a developing country.

Several hospital-based studies have shown the beneficial effect of kangaroo care on preterm infants. Long-term outcome was studied in 297 preterm infants born at Harare Hospital weighing 500-1800 g, discharged home on kangaroo care and followed up for 12 months. Of these, 79 (26.6%) died, 141 (47.5%) survived to complete follow-up and 77 (25.9%) were lost to follow-up. Of those who died, median birthweight was 1460 g, median age at hospital discharge 7 days, median weight at discharge 1400 g and median age at death 66 days. Of those who completed follow-up, median birthweight was 1575 g, median age at hospital discharge was 6 days and median weight at hospital discharge was 1500 g. Of those who were lost to follow-up, median age at loss to follow-up was 70 days, median birthweight was 1540 g, median age at hospital discharge was 5 days and median weight at hospital discharge was 1500 g. The hospital re-admission rate was 22.9% with 8.8% mortality. Maternal mortality and chronic morbidity rates were 4.7% and 7.4%, respectively. On comparing those who died with those who completed follow-up, mother's age <20 years, birthweight <1500 g and maternal mortality and chronic morbidity were significant risk factors for infant mortality. Age at discharge and weight at birth and on discharge were not significantly associated with infant mortality.

Caregivers' perceptions and experiences of 'kangaroo care' in a developing country.

The widespread use of 'kangaroo care' is yet to be realized despite strong evidence to suggest that this method of preterm care is safe, effective and affordable. We need to understand users' perception of this method of care. We studied, through focus group discussions, caregivers' experiences and perceptions of this method in a tertiary level hospital of a developing country. We conclude that, in this hospital, caregivers preferred kangaroo care to conventional methods. Communities' awareness of this method of care and its advantages must be improved.

Levels and risk factors for mortality in infants with birth weights between 500 and 1,800 grams in a developing country: a hospital based study.

OBJECTIVE: To determine levels of mortality and risk factors for mortality in infants born with birth weights below 1,800 gms. DESIGN: Prospective descriptive study. SETTING: Harare Central Hospital in Zimbabwe. SUBJECTS: All infants born and admitted to Harare Hospital Neonatal Unit between January and May 2000, with birth weight between 500 gms and 1,800 gms. STUDY FACTORS: Mothers' age, parity, booking status, mode of delivery, infants' sex, birth weight, use of intensive care, outcome in hospital, age at death and age at discharge from hospital. RESULTS: Four hundred and ninety infants were studied. In hospital, the fatality rate was 39.4%. Only 49.4% of mothers had received antenatal care. The median birth weight of those who died was 1,077 gms (Q1 = 500, Q3 = 1,357) while that of the discharged infants was 1,530 gms (Q1 = 850, Q3 = 1690). Risk factors for mortality were birth weight less than 1,500 gms compared to 1,500 gms or more. Odds Ratio (OR) 7.53 (95% CI = 4.66 to 12.23), breech delivery compared to vaginal delivery, OR 2.40 (95% CI = 1.28 to 4.52) and lack of antenatal care OR 1.59 (95% CI = 1.08 to 2.33). Parity, sex of infant and receiving intensive neonatal care were not significantly associated with mortality. CONCLUSION: Strategies to reduce mortality in these infants should include better access to early, high quality obstetric care and avoidance of breech delivery in preterm infants. Identification of avoidable factors leading to preterm delivery is critical.

Perinatal practices in two rural districts of Zimbabwe: a community perspective.

OBJECTIVE: To describe perinatal practices from a community perspective and identify factors associated with perinatal death. DESIGN: Cross sectional community based survey. SETTING: Murewa and Madziwa rural areas, Zimbabwe. SUBJECTS: Women aged 15 to 50 years who had been pregnant within the 24 months preceding the survey. MAIN OUTCOME MEASURES: Where delivered, where preferred to deliver, model of delivery, use of herbs in labour, duration of labour, assistant at delivery, time of delivery, condition of baby at birth, resuscitation methods, birth weight, initiation of breast feeding, illness in the first week and outcome of pregnancy. RESULTS: 644 women were interviewed; 581/644 stated where they would have liked to deliver and 505/644 stated where they actually delivered their last baby. The majority 369/581 (62.4%) preferred to delivery at a government hospital and 240/505 (47.5%) actually delivered at a government hospital. Of the home deliveries only 27/581 (4.6%) preferred to deliver at home and yet 123/505 (24.4%) actually delivered at home. Primary care clinics were less preferred 151/581 (25.5%) as a place for delivery and 89/505 (17.6%) actually delivered there. Labour lasting more than 12 hours occurred in 20.4% of deliveries. Nurses were the commonest attendants at delivery 309/508 (60.4%) and morbidity following delivery was noted in 72/495 (14.5%). Resuscitation was carried out in 61/72 infants. Beating/shaking 36/61 (58.0%) and pouring cold water over the baby 11/61 (18.0%) were the commonest methods of resuscitation. Being delivered by a doctor compared to a nurse and being in Murewa district were statistically significant risk factors for mortality with Odds Ratio (OR) 5.21 (95% CI 2.86 to 9.51) and 3.90 (95% CI 1.51 to 10.09) respectively. The odds of dying when delivered by breech extraction were high, but not statistically significant OR 3.73 (95% CI 0.92 to 13.97) when compared to being delivered by vertex delivery. Labour more than 12 hours, use of herbs in pregnancy and time of delivery were not significantly associated with mortality with OR (95% CI) of 1.02 (0.40 to 2.19), 0.92 (0.00 to 4.38), 1.05 (0.56 to 1.97) respectively. On logistic regression analysis only being delivered in Murewa district remained significant. CONCLUSION: The utilisation of primary health care centres for delivery was unexpectedly low and home deliveries were unacceptably high. Increased mortality when delivered by a doctor and high early neonatal morbidity suggest poor monitoring and delayed intervention in labour. Infant morbidity following delivery was high and methods for neonatal resuscitation inappropriate. There is a need for more studies looking into health worker skills particularly in the areas of partogram use and neonatal resuscitation in these districts.

Situation analysis of obstetric care services in a rural district in Zimbabwe.

OBJECTIVE: To conduct a situation analysis of obstetric services in a rural district of Zimbabwe. DESIGN: Observational study. SETTING: 13 primary health care centres in Murewa district in Zimbabwe. MAIN OUTCOME MEASURES: Number of maternity beds, antenatal attendance, deliveries per month, availability of antenatal, intrapartum and neonatal care equipment, intrapartum monitoring and neonatal resuscitation skills. RESULTS: 13 of 15 primary health care clinics providing obstetric care in Murewa district were surveyed in 1995. Median number of maternity beds were nine (Q1 = 0, Q3 = 11) per clinic, median number of first ANC attenders per month was 15 (Q1 = 3, Q3 = 18), median number of deliveries per clinic per month were eight (Q1 = 0, Q3 = 16). While all clinics had laboratory facilities, 6/13 could estimate haemoglobin, 5/13 syphilis serology, none of the clinics sent blood to district hospitals for blood grouping and there were no microscopes at clinics for malaria parasite determination. Only 6/13 clinics used partographs for monitoring labour, 10/13 had suction machines for neonatal resuscitation while only 3/13 had ambu bags, 3/13 had oxygen and 2/13 had heaters. Correct methods for neonatal resuscitation were used in 3/13 clinics. Clean water supply, reliable power supply and the referral system were not optimal. CONCLUSION: Basic equipment for antenatal, intrapartum and neonatal care was inadequate. Essential laboratory facilities for obstetric care were lacking. Skills for intrapartum monitoring and neonatal resuscitation were inadequate. The referral system was poor. There is need for more strategic planning at primary health care level in this district which is known to have high perinatal and neonatal death rates. More emphasis should be placed on strengthening basic laboratory back up service for obstetric care, strengthening infrastructural and referral systems as well as training in areas of lost or no skills.

Kangaroo care for well low birth weight infants at Harare Central Hospital Maternity Unit--Zimbabwe.

OBJECTIVE: To describe the experience in a newly established Kangaroo Care Unit (KCU) at a tertiary level hospital and to identify factors associated with poor outcome in this unit. DESIGN: Cross sectional study. SETTING: Kangaroo Care Unit at Harare Central Hospital, Zimbabwe. SUBJECTS: Mothers admitted to the KCU and their well preterm infants. MAIN OUTCOME MEASURES: Discharge home or referral back to the Neonatal Unit (NNU) for conventional care. RESULTS: 613 mother infant pairs were studied from May 1994 to December 1996. The median age for all mothers was 23 years (Q1 = 15, Q3 = 26). Fifty four percent of the infants were female. Median age at admission to KCU was 12 days (Q1 = 1, Q3 = 25). Seventy two percent of infants were discharged home from the KCU. The rest (28%) were referred back to NNU for conventional care. The odds of being referred back to the NNU were significantly higher if the infant was male OR = 1.82 (95% CI: 1.25 to 2.66); if the birth weight was < 1 500 gms OR = 1.52 (95% CI: 1.04 to 2.22); if the admission weight to the KCU was < 1500 grams OR = 2.16 (95% CI: 1.42 to 3.29) or if the age on admission to KCU was 14 days or more OR = 2.15 (95% CI: 1.44 to 3.29). These factors remained significant after adjusting for confounding. Mother's age, parity, booking status or whether admission was during the cold months or not had no significant bearing on the outcome in this unit. Reasons for referral back to NNU included apnoea (27%); respiratory distress (27%); aspiration pneumonia (18%); neonatal jaundice (8%); poor feeding (7%); ill mother (5%); sepsis (4%) and diarrhoea (3%). On multivariate analysis birth weight was the strongest predictor for being referred back to the NNU OR = 10.753 (95% CI: 6.026-19.186). CONCLUSION: Establishment of a KCU at a tertiary level hospital is feasible. Kangaroo care for well preterm infants is suitable for most mothers and their preterm infants. Infants were more likely to be referred back for conventional care if they were male, very low birth weight and if the age at admission was greater than two weeks. Further studies are needed to determine the long term survival of these infants.

Antenatal care patterns and factors associated with perinatal outcome in two rural districts in Zimbabwe.

OBJECTIVE: To study antenatal care (ANC) patterns, to identify factors associated with poor perinatal outcome and quality of ANC. DESIGN: Cross sectional community based survey. SETTING: Murewa and Madziwa rural Districts. SUBJECTS: 644 women aged between 15 to 50 years who had been pregnant in the immediate 24 months preceding the survey. MAIN OUTCOME MEASURES: Perinatal outcome, ANC attendance, distance from health centre, ability to pay for ANC, gestation at first booking, frequency of ANC visits, use of herbs in pregnancy, pregnancy related morbidity and quality of ANC. RESULTS: A total of 644 women were interviewed. Overall perinatal mortality was 115 per 1,000 births. 511/644 (79.3%) visited a health centre for ANC. The woman herself (41.8%) as well as husbands (41.8%) were commonly the decision makers regarding starting ANC. Only 298/510 (58.4%) of women used primary health care facilities for ANC. 211/629 (35.1%) of women lived more than 5 kms from the health care centre. 153/629 (24.3%) were not able to pay for ANC. Only 110/509 (21.6%) started ANC in the first trimester. 307/495 (62.0%) made five or less ANC visits. Pregnancy related morbidity was high 209/644 (32.5%) and use of traditional herbs was common 158/644 (24.%). There were significant differences between districts as far as perinatal outcome, ANC attendance, distance from health centre, ability to pay for ANC, gestation at first booking, frequency of ANC visits, use of herbs in pregnancy, pregnancy related morbidity and quality of care all having p values of 0.001 or less. On further analysis using logistic regression, having problems with pregnancy and vaginal bleeding were significant predictors of poor perinatal outcome with odds ratio (95% CI) of 2.8 (1.4 to 5.9) and 3.0 (1.1 to 8.6) respectively. Quality of ANC was rated as sub-optimal. CONCLUSION: Majority of rural women attended clinics for ANC. Perinatal mortality rate and pregnancy related morbidity were high. Vaginal bleeding was the strongest predictor for mortality. There is need to improve quality of antenatal care as this was found to be suboptimal. More objective oriented antenatal care visits and waiting mother's shelters form part of the solutions. Further studies are called for.

Aero-allergen sensitisation patterns amongst atopic Zimbabwean children.

OBJECTIVE: To characterize children presenting with atopic conditions using the RAST test. DESIGN: Retrospective descriptive study. SETTING: General paediatric clinic in the private sector. SUBJECTS: 84 children aged below 12 years, who had the RAST test, who presented to a general paediatric clinic between 1993 and 1998 with atopic conditions for care. RESULTS: The median age for all children in the study was 52 months. Forty eight were male and 36 female. Eczema (33.9%) was the most frequent clinical diagnosis especially in those less than 24 months of age, followed by asthma (25.5%), allergic conjunctivitis (24.0%) and allergic rhinitis (15.6%). Total IgE was not statistically significantly associated with clinical diagnosis(p = 0.889), age of the child (p = 0.102), gender (p = 0.687) or absolute eosinophil count (p = 0.318). The commonest allergens identified were dust mite (Dermatophygoides pteronissinus and D. farinae) and Bermuda grass. While antibody reaction to weeds, particularly plantain, were also common, these reactions were mostly mild to moderate. Allergy to cats and moulds was rare. CONCLUSION: In the absence of routine testing for specific allergens avoidance of dust mite and Bermuda grass seem important strategies in the management of difficult children with atopy. There is need for a prospective study to shed more light on the allergens that cause these common atopic conditions in our environment.

Risk factors for neonatal mortality: Harare Central Hospital Neonatal Unit--Zimbabwe.

OBJECTIVE: To assess risk factors for neonatal mortality in a tertiary level neonatal unit. DESIGN: Case control analysis of routine neonatal data for 1998. SETTING: Harare Central Hospital Neonatal Unit. SUBJECTS: All neonates delivered at Harare Maternity Hospital and admitted to the neonatal unit for care between January and December 1998. MAIN OUTCOME MEASURE: Neonatal mortality in hospital. RESULTS: A total of 5,305 neonatal admissions were studied of which 19.3% died in hospital. The median age at death was two days (Q1 = 1, Q3 = 3) and the median age at hospital discharge was 3 days (Q1 = 1, Q3 = 6). Risk factors for mortality were un-booked mother odds ratio (OR) 2.36 (95% CI = 1.98 to 2.81), breech delivery OR: 1.76 (95% CI = 1.39 to 2.22), low birth weight OR: 4.67 (95% CI = 3.92 to 5.57), prematurity OR: 2.36 (95% CI = 2.09 to 2.66), congenital malformations OR: 2.80 (95% CI = 1.72 to 4.53) and birth asphyxia OR: 1.79 (95% CI = 1.51-2.12). Being admitted for respiratory distress was associated with better survival OR: 0.22 (95% CI = 0.17 to 0.28). Having a Caesarian section was also protective OR: 0.60 (95% CI = 0.47 to 0.76). Mother's age, parity, time of delivery and sex were not significantly associated with mortality odds ratios (95% CI) of 1.07 (0.86 to 1.34), 0.94 (0.78 to 1.13), 1.10 (0.93 to 1.30) and 0.89 (0.78 to 1.03) respectively. On regression analysis birth weight greater than 2,500 g, being un booked and breech delivery were predictive of mortality with OR (95% CI) of 0.99 (0.99 to 0.99), 1.31 (1.12 to 1.61) and 1.15 (1.04 to 1.28) respectively. CONCLUSION: Low birth weight is the highest risk factor for mortality in this tertiary level hospital. Strategies targeted at low birth weight infants such as antenatal corticosteroid use, improved intrapartum care, appropriate antibiotic use, improved efficiency and access to neonatal intensive care will have the most impact on neonatal mortality.

Syphilis in Murewa District, Zimbabwe: an old problem that rages on.

OBJECTIVE: To determine the prevalence of syphilis among pregnant women and women giving birth in health centres in a rural district and to identify problems associate with syphilis control in the same district. DESIGN: Cross sectional descriptive study. SETTING: Murewa District health facilities. SUBJECTS: Women attending health facilities in this district for antenatal care or delivery between February and May 1993. MAIN OUTCOME MEASURES: Syphilis sero-prevalence rate. Factors associated with poor syphilis control. RESULTS: Even though it is recommended that all women attending clinics for antenatal care (ANC) should be screened for syphilis at first visit only 308 (20%) out of 1,556 first visit attenders were screened during the study period. Three hundred and sixty six (33%) out of 1,096 women giving birth in health institutions were screened. The RPR/TPHA sero positivity rate for antenatal women was 9.2% while that for women delivering was 9.8%. A positive RPR was not significantly associated with the women's age, parity, infant's birth weight, sex or pregnancy outcome. Factors associated with poor syphilis control in this district included: lack of motivation and appreciation of the seriousness of syphilis in pregnancy; lack of transport to send specimens and receive results from Murewa District Hospital; poor record keeping; loss to follow up of women being tested or after starting treatment; lack of contact tracing and treatment of contacts and difficulties in implementing the 10 day neonatal regime and follow up of these infants. CONCLUSION: Syphilis remains poorly controlled in Murewa district and may be contributing significantly to high perinatal mortality rates. There is need to strengthen the syphilis control programme through motivation and training of health workers, decentralisation of testing and treatment of the condition and improved contact tracing. A repeat RPR test at delivery may not be cost effective.

PIP: A cross-sectional descriptive study was conducted at Murewa District health facilities to determine the prevalence of syphilis among antenatal women and women giving birth and to identify problems associated with syphilis control. The data were collected from the health center nursing staff who took care of the women during antenatal period or delivery at Murewa District health facilities. The results showed that only 308 (20%) out of 1556 first visit attendees were screened during the study period while only 366 (33%) women giving birth were screened out of 1096 subjects. The RPR/TPHA seropositivity rate for antenatal women was 9.2%, while that for women delivering was 9.8%. There was no significant association between positive RPR result and the women's age, parity, the infant's birth weight or sex or pregnancy outcome. Poor syphilis control in Murewa District Health facilities was due to lack of motivation and appreciation of the seriousness of syphilis in pregnancy, lack of transport to send specimens and receive results from Murewa District Hospital, poor record keeping, loss during follow-up of women during testing or after treatment, lack of contact tracing and treatment of contacts and difficulties in implementing the 10 day neonatal regimen and follow up for these infants. Owing to these factors, syphilis remains poorly controlled in Murewa district and may be considered a significant contributing factor to high perinatal mortality rates. Action is needed to strengthen the syphilis control program through motivation and training of health workers, improved contact tracing, and decentralization of testing and treatment.

Kangaroo care versus incubator care in the management of well preterm infants--a pilot study.

This pilot study was conducted to compare the effectiveness of the kangaroo care method with current, mainly incubator-based care in managing well preterm infants in a tertiary level hospital in a developing country. Altogether, 74 infants (37 per group) were consecutively allocated to receive either kangaroo care or incubator care. After adjusting for age and weight on admission to the study, we found that infants in the kangaroo care group gained twice as much weight per day (20.8 vs 10.2 g, p = 0.0001), had a shorter stay in hospital (16.6 vs 20.7 days, p = 0.0457) and had a better survival rate (0% vs 9% deaths). Also, they were ill less frequently, but after adjusting for age and weight this difference was not significant. This pilot study suggests that the kangaroo care method has major advantages over incubator care of preterm infants in our hospital. Hospitals which cannot use incubators optimally may find kangaroo care to be a better method of improving perinatal and neonatal morbidity and mortality.

Perinatal mortality rates and associated socio-demographic factors in two rural districts in Zimbabwe.

OBJECTIVE: To determine perinatal mortality rates and associated socio-demographic factors in two rural districts in Zimbabwe. DESIGN: Cross sectional community based survey. SETTING: Murewa and Madziwa rural districts. SUBJECTS: Women aged 15 to 50 years who had been pregnant over the preceding two years before the study. MATERIALS AND METHODS: A questionnaire was administered to eligible women by trained interviewers. Information pertaining to the women's socio-demographic characteristics, reproductive health profile and pregnancy outcome was documented. A post hoc case control analysis was undertaken to determine the risk factors associated with poor perinatal outcome. Women who had a poor perinatal outcome were designated cases and those with a good outcome were designated controls. RESULTS: The average perinatal mortality rate for both districts was 111 per 1,000 live births (Murewa 182 per 1,000 and Madziwa 48 per 1,000). Factors significantly associated with perinatal mortality were ethnicity, marital status, subjective standard of living and the women's level of formal education. Being Zezuru or being married was associated with poor perinatal outcome. Living well and having high levels of education were also associated with poor perinatal outcome. Perinatal mortality was not significantly associated with maternal age or spouse level of education. CONCLUSION: This study showed unacceptability high perinatal mortality rates in these rural districts. The true socio-demographic factors associated with perinatal mortality could not be ascertained in this study because of confounding factors. There is need to study quality of antenatal, intrapartum and neonatal care offered by health centres in these districts. In addition there is need to strongly advocate a perinatal programme to address these high mortality rates.

PIP: A community-based survey of 640 women 15-50 years old who had been pregnant in the 2 years preceding the study was conducted in two rural districts in Zimbabwe (Murewa and Madziwa) to assess perinatal mortality and the associated sociodemographic factors. The overall perinatal mortality rate was 111/1000 live births (182/1000 in Murewa and 48/1000 in Madziwa). Murewa had significantly more stillbirths, abortions, and perinatal and infants deaths than Madziwa. A perinatal death was 1.94 times more likely if the woman was from the Zezuru ethnic group. Single, separated, divorced, and widowed women had better pregnancy outcomes than married women (polygamous or monogamous). Finally, living well (subjectively assessed) and a high educational level (completion of primary school and above) were both associated with significant risks of poor perinatal outcomes. Maternal age and inter-birth interval were not risk factors. Because of lack of controls for confounding, the generally puzzling risk factors identified in this study require confirmation. However, the high mortality rate identified in this study indicates an urgent need for implementation of effective perinatal programs in this area.

The prevalence of nipple disease among breast feeding mothers of HIV seropositive infants.

OBJECTIVE: To determine the prevalence of nipple disease among breast feeding mothers of symptomatic HIV seropositive infants and factors associated with nipple disease. DESIGN: Cross sectional survey. SETTING: Harare Central Hospital general paediatric wards. SUBJECTS: One hundred and four symptomatic, HIV seropositive breast feeding infants and their mothers. MAIN OUTCOME MEASURES: Prevalence of nipple disease. RESULTS: The majority of the hospital admissions (90%) were for pneumonia. The prevalence of nipple disease was high (30.8%). Nipple eczema was seen in 22.1%, cracked nipples in 10.6% and sore nipples in 10.6% of these breast feeding mothers. The odds of developing nipple disease in the mother if the infant had oral disease were 11.47 (95% CI 5.28 to 25.39). There was no significant association between nipple disease and mother's age, infant's age, nutrition status or mode of feeding. Malnutrition was a major problem. CONCLUSION: Nipple disease was highly prevalent and oral disease was the major risk factor for the development of nipple disease in breast feeding HIV seropositive mothers.

Ability of caregivers to recognise signs of pneumonia in coughing children aged below five years.

OBJECTIVES: This survey was carried out to determine: 1. The ability of caregivers to recognise signs of pneumonia in children aged below five years who are coughing. 2. The proportion of caregivers who possess a watch and are able to use that watch to count respiratory rates in children. 3. The home remedies used for cough. DESIGN: A cross sectional study. SETTING: Three Primary Health Care Clinics and a tertiary level hospital in Harare. MATERIALS AND METHODS: 413 children aged less than five years attending a health care centre for cough were studied. Socio-demographic and personal characteristics of both mother and child were documented. An inventory of home remedies that were being used since onset of cough was taken. The child was also examined for signs of pneumonia by both caregiver and research nurse. The sensitivity and specificity of the caregiver's ability to recognise signs of pneumonia compared to the research nurse was calculated. RESULTS: The ability of caregivers to recognise signs of pneumonia was high (sensitivity 95pc for hospitalized cases and 85pc for the clinic cases). Sensitivity tended to drop with increase in child's age. The caregiver's ability to recognise signs of pneumonia was not significantly related to the number of children she had or her level of education. Only 19.5pc of caregivers owned a watch and could use a watch to count respiratory rates correctly. Those who owned a watch were more likely to count respiratory rates correctly. Those who claimed ability to use a watch actually counted respiratory rates correctly. Sixty six pc had used some remedy to treat the cough at home. CONCLUSIONS: Caregivers were able to recognise signs of pneumonia without a watch. Only a small proportion of caregivers own watches and are able to use them correctly to count respiratory rates. National ARI control programmes should emphasize caregivers observing breathing when a child has a cough and encourage seeking early and appropriate health care when pneumonia is detected. Watches or timers should be made available and be recommended for use by trained health workers in health centres only. Use of potentially harmful remedies should be discouraged.

Measles epidemic in Harare, Zimbabwe, despite high measles immunization coverage rates.

Despite rapidly increasing measles immunization coverage in Harare city, measles remains endemic, and regular outbreaks occur. The most recent occurred in 1988, when the measles immunization coverage was 83%. We have carried out a retrospective study of the clinical and epidemiological features of this outbreak to assess whether the present immunization policy needs to be changed. Of 4357 cases of measles seen at primary health care centres and hospitals in Harare during the outbreak, 1399 (32%) were severe or involved complications that required hospital admission. The peak incidence occurred among under-2-year-olds, followed by that among 5-7-year-olds. Poor nutritional status was significantly more frequent among children who were hospitalized and among those who died. A total of 59% of all cases aged 9-59 months had documented evidence of measles immunization. The most frequent complications, which occurred most often among under-5-year-olds, were diarrhoea with dehydration, pneumonia, laryngotracheobronchitis, and convulsions, which together affected 56% of hospitalized cases. The hospital case fatality rate was low (1.43%). In Harare, measles transmission remains a problem, despite high measles immunization coverage rates; the failure rate for the standard Schwarz measles vaccine also appears to be high. There is a need to reduce the number of measles cases among under-9-month-olds and young children. Further studies into alternative measles vaccines and schedules are required.

Measles epidemic in Harare, Zimbabwe, despite high measles immunization coverage rates

Despite rapidly increasing measles immunization coverage in Harare city, measles remains endemic, and regular outbreaks occur. The most recent occurred in 1988, when the measles immunization coverage was 83%. We have carried out a retrospective study of the clinical and epidemiological features of this outbreak to assess whether the present immunization policy needs to be changed. Of 4357 cases of measles seen at primary health care centres and hospitals in Harare during the outbreak, 1399 (32%) were severe or involved complications that required hospital admission. The peak incidence occurred among under-2-year-olds, followed by that among 5-7-year-olds. Poor nutritional status was significantly more frequent among children who were hospitalized and among those who died. A total of 59% of all cases aged 9-59 months had documented evidence of measles immunization. The most frequent complications, which occurred most often among under-5-year-olds, were diarrhoea with dehydration, pneumonia, laryngotracheobronchitis, and convulsions, which together affected 56% of hospitalized cases. The hospital case fatality rate was low (1.43%). In Harare, measles transmission remains a problem, despite high measles immunization coverage rates; the failure rate for the standard Schwarz measles vaccine also appears to be high. There is a need to reduce the number of measles cases among under-9-month-olds and young children. Further studies into alternative measles vaccines and schedules are required