In this month's Bulletin

Self-help health insurance works (pp. 613–621)

Where a system of national risk-pooling is not available, community-based health insurance can provide effective protection against being ruined by unexpected health care costs. The Medical Insurance Fund of the Self Employed Women's Association in Gujarat was able to reimburse 1712 out of 1927 claims over a six-year period, 47% of them in full, and 53% in part (at a mean reimbursement rate of 55.6%). Reimbursement more than halved the percentage of catastrophic hospitalizations, i.e. those costing more than 10% of the annual household income and resulting in impoverishment. The scheme was set up in 1992 by women without a regular income. The annual premium for life insurance, medical insurance and asset insurance is 72.5 rupees (US$ 1.67), of which 30 rupees are earmarked for medical insurance. This provides maximum coverage of 1200 rupees (US$ 28). (See also Editorial, p. 612.)

Epidemiological estimates: comparing calculated and empirical data (pp. 622–628)

Empirical data are often incomplete; for instance deaths from a particular disease are more likely to be recorded than each case of that disease. The missing data can be supplied to some extent by back-calculation. To test the accuracy of models for carrying out such procedures, two internally consistent artificial data sets are compared to four high-quality empirical data sets. Large discrepancies are found between the calculations and the measurements. Inaccurate data and variations in past trends may be the cause. (See also Editorial, p. 611.)

No association found between female circumcision and perinatal death (pp. 629–632)

High perinatal mortality among the children of African women in Sweden raised the question of a link between these deaths and female circumcision. However, a study of 63 perinatal deaths of infants born to circumcised women in Sweden between 1990 and 1996 produced no evidence of such an association. Other factors, such as maternal behaviour, misunderstandings, and insufficient obstetric care, appear to have contributed to these deaths.

Urinary iodine: median concentrations of 100–199 µg/l are sufficient (pp. 633–636)

The concentration of iodine in the urine is the prime indicator of a person's nutritional iodine status. Median urinary iodine concentrations of 100–199 µg/l are thought to indicate adequate iodine intake and optimal iodine nutrition. To answer the question of whether this threshold should be raised in order to reduce the frequency of iodine deficiency (defined as concentrations of under 50 µg/l), data from 48 populations were analysed. The data showed that in populations with a median of over 100 µg/l the percentage with under 50 µg/l was 4.8% — much lower than the 20% assumed previously. The authors onclude that the threshold does not need to be raised.

Diarrhoeal disease reduction in the Philippines (pp. 637–643)

Deaths caused by diarrhoea were reduced by about 5% annually over an 18-year period. Hospital admission rates for diarrhoea went down by about 2.4% a year. The National Control of Diarrhoeal Disease Programme played an important part in achieving this improvement by organizing access to oral rehydration therapy. Other factors contributing to it are increased breastfeeding, improved nutrition, safer water and better sanitation.

How comparable are DALY figures cross-nationally? (pp. 644–652)

Variation in disability-adjusted life-year (DALY) estimates in Denmark, England and Wales, France, the Netherlands, Spain and Sweden is examined, using dementia in women as an example. Conclusion: summary measures of population health are only as reliable as the weakest link in the chain, which is the epidemiological evidence. This attempt to estimate the burden of dementia in a cross-nationally comparable manner shed light on some of the gaps in the available data. The findings will help to guide agenda-setting for the collection of epidemiological data.

The cost of malaria control in China (pp. 653–659)

The authors examine the costs of vector surveillance, population blood surveys and case management in Henan Province for one year, during which 12 325 suspected cases of malaria were reported. They conclude that government spending should continue at at least its current level of US$ 0.03 per person at risk. Further cuts in current spending would increase future costs by allowing incidence to reach epidemic proportions.

A complete system for local malaria control (pp. 660–666)

Starting in 1994, the following interventions were made in a commune of 716 inhabitants in southern Viet Nam: distribution of insecticide-treated bednets to all households, re-treated every six months; a health post with staff trained in malaria microscopy and treatment; a regular supply of malaria drugs and other essential control materials; annual parasitaemia surveys with treatment of infected subjects; and a community involvement programme. By 1999 the proportion of parasitaemic subjects had dropped from 42% to 4%. The approach is effective and could be used in a variety of situations.

Linking health sector reform to reproductive health needs (pp. 667–674)

The agendas for reproductive health and health sector reform are both evolving rapidly but too often without reference to each other, making efforts in both areas disjointed. Dialogue and collaboration should be promoted between the policy groups concerned. They need to build up a common understanding on relevant issues, agree on aims, link decisions on structural change to reproductive health evidence, and strengthen the framework and skills needed for policy-making.

Basing decisions on needs in Windhoek, Namibia (pp. 675–681)

A review of primary health care services revealed disparities between the utilization of services and the allocation of staff. Poorer areas were less provided for although their needs were greater. Centrally made decisions on resource allocation had reinforced the disparity. The study resulted in a redistribution of staff and a strong case for allocating capital expenditures more effectively.