Estimating the influence of health as a risk factor on unemployment: a survival analysis of employment durations for workers surveyed in the German Socio-Economic Panel (1984-1990).

In this paper we examine the link between unemployment and health. The negative health selection hypothesis, which proposes that poor health poses an unemployment risk, is tested using data from the German Socio-Economic Panel (GSOEP). The statistical influence of health related variables on the duration of employment for a cohort of workers is estimated. Results from the Cox proportional hazards regression model show gender and nationality specific negative selection. In the event of a long or chronic illness female workers are at a higher risk of unemployment than male workers. Whereas chronic illness raises the probability of unemployment among foreign workers, there is no statistical evidence for this for German workers. The paper, thus, shows that health factors determining unemployment affect different types of workers in different ways. Consequently, results from aggregate studies may be misleading. A second result of the paper is that, irrespective of gender and nationality, there is strong evidence for lagged state dependence on previous spells of unemployment, i.e. individuals who had experienced unemployment previously were more at risk of renewed unemployment than those without such spells. These findings do not only confirm the selection hypothesis, but also illustrate how labour market risks are closely associated with attributes of social inequality and how this could result in the accumulation of risks for those who are socially or politically vulnerable in the labour market.

Estimating child mortality in Zimbabwe: results of a pilot study using the preceding births technique.

Using the preceding birth technique, 2,229 mothers were interviewed at four antenatal centres (two of which were urban) over a three month period during a pilot study in Zimbabwe. Results of the study showed that there was a small difference between the under two and under five mortality. The smaller than expected difference could have been due to some degree of urbanisation being experienced at two of the centres, and to a systematic selection bias of the method. Other findings of the study showed that younger mothers (under 20 years of age) and older mothers (over 40 years) experienced higher proportions of mortality, than mothers in the age group between. The higher proportion of male mortality compared to that for females reflected the expected trend, but the size of the difference was somewhat surprising, and could have been due to recall bias. The mean birth interval was 36 months, rather than the expected 30 months, and analysis of mortality in relation to birth interval and maternal age showed that a birth interval of less than 18 months was associated with higher child mortality, significantly so with the second last child, independent of mother's age. Longer birth intervals (more than three years) among older mothers were associated with higher changes of child survival. It is concluded that the Brass-Macrae method is a useful technique to analyse levels and trends of child mortality.

PIP: The University of Zimbabwe and the Zimbabwe Congress of Trade Unions tested the ability of the Brass-Macrae methodology (i.e., preceding birth technique) to estimate current levels and trends in child mortality. Interviews were conducted with 2229 mothers attending the prenatal clinics of the Mbare and Glen View polyclinics in Harare and of a provincial hospital (Marondera) and district hospital (Mutoko), both in Mashonaland East Province. The overall child mortality rate was 53/1000 of the last live births and 54/1000 of the second last live births. Some degree of urbanization and a systematic selection bias of the method may have contributed to the small difference between the two rates. Among last live births, the provincial hospital in Marondera had the lowest child mortality rate, while the Mbare polyclinic in the capital city of Harare had the highest (43 vs. 61). The difference in child mortality rates for last live births between urban and rural areas was slight, while it was much greater for second last live births (2 vs. 11 points). Among second last live births, the district hospital in Mutoko had the lowest rate, while the Marondera provincial hospital had the highest rate (39 vs. 64). Child mortality levels peaked in the mothers' age groups less than 20 (88 for last live births) and more than 40 (84 for last live births and 78 for second last live births). Among last and second last live births, boys were more likely to die than girls (64 vs. 42, p 0.05, and 64 vs. 45, p 0.05, respectively). This difference could have been due to boys being more frail than girls or mothers being able to better recall boys deaths, since boys are highly valued. The mean birth interval was 36 months, but the expected interval was 30 months. A birth interval of less than 18 months was significantly associated with higher child mortality among second last live births (p 0.001), regardless of mother's age. Birth intervals of more than 3 years were associated with the best child survival rates in every maternal age group (e.g., 30-34 years, 8 vs. 52-250). This pilot study showed that the Brass-Macrae method is useful to estimate child mortality.

Value of the history and stool occult blood test in selection of patients for upper endoscopy in Zimbabwe.

Patients with abdominal pain and no definite diagnosis referred for endoscopy were studied to define discriminating features in the history, and the value of a stool occult blood test, in predicting the presence of upper gastrointestinal disease. Endoscopy was performed in 116 patients; pathology was seen in 32 (duodenal ulcer 17, gastric carcinoma 4, gastric ulcer 3, miscellaneous 8) and no pathology was seen in 84 patients. Features that predicted upper gastrointestinal pathology were, in descending order of rank: a positive pointing sign, a positive stool Fecult test, a history of vomiting, loss of weight, and alcohol intake. Using these discriminating features together it was possible to correctly predict 95% of patients with abnormal endoscopy and 82% of patients with a normal endoscopy. The history and the stool occult blood test are useful predictors of the presence of upper gastrointestinal pathology and may aid rational selection of patients for endoscopy.

Variation in cuff blood pressure in untreated urban outpatients with mild hypertension in Harare, Zimbabwe--implications for the diagnosis of hypertension.

Duplicate measurements of BP were recorded at six visits in 32 male urban factory workers found to have raised BP (systolic blood pressure (SBP) greater than or equal to 160 mmHg or diastolic blood pressure (DBP) greater than or equal to 90 mmHg) at the initial visit). A systematic decrease in both SBP and DBP occurred over the first three visits. The BP at visits 4-6 was not significantly different. Most of the decrease in BP occurred between the first and second visit (mean SBP decreased 8.5 mmHg and mean DBP decreased 11.1 mmHg). After 6 visits 66% of subjects could be classified into clear clinical groups (normotensive 53.5%, hypertensive 12.5%) but in 34% the variation in cuff BP precluded classification. In these subjects prolonged observation or home BP monitoring may be required before initiating lifelong treatment.