A population health study from Iceland published this month† examined maternity outcomes according to area of residence and local service model. Dr. Sigridur Haraldsdottir and his colleagues conducted a study of nearly 41,000 singleton births between 2000 and 2009 with particular attention to the role of geography on health.
Iceland has some unique geographic health service features. Metro Reykjavik holds over 200k of the country’s 325k people. The largest city outside the capital region holds just over 18k people (a few more people than Salmon Arm and a few less than Cranbrook, for perspective). Of Iceland’s 13,000 kilometres of roads, roughly 64% are unpaved and nearly 63% of the landmass is tundra.
In that context, the authors defined geographic zones by distance to the capital region – within 70km, between 70-250km, and more than 250km. Though geography was not associated with more pre-term or low birth weight babies, it was associated with higher rates of perinatal mortality for some specific groups. For instance, higher rates of perinatal mortality were found in the period 2005-2009 for women from rural areas who delivered in the highest tier maternity units (which are urban based), indicating that there may be associated factors in the referral and travel itself that impacts outcomes for high-risk moms from rural areas. Irrespective of birth location, the highest risk categories were residents in the peri-urban category (within 70km of Capital region) and those furthest flung (greater than 250km), with a non-significant outcomes advantage for those in the middle group.
The service level of ‘some but limited specialist services’ was also more likely to record perinatal mortality as compared to the capital region reference group of ‘diverse specialist services.’ But it was the service category of ‘no health services’ that saw the greatest disadvantage. Those areas with generalist only services did not have the same outcomes disadvantage, although the parsing of the data began to undermine the initially strong sample size and so statistical significance was not found.
That pattern fits with data we are seeing from around the world under seemingly different conditions of centralization. For BC residents with no local birth services, the outcomes are dangerously worse. Pregnant women who live one to two hours from services are more than six-times more likely to have an accidental, out-of-hospital birth, and those four or more hours from services are three times more likely to experience a perinatal death.‡
The conclusions of Haraldsdottir et al pertain mostly to the effectiveness of the referral system in rural areas, citing a formalization of referral policy in 2007 and lower recorded rates of gestational diabetes and hypertension as evidence of lower diagnostic activity or under-reporting of risk factors. Further study may be needed to tease out why generalist services and those 70-250km from the capital region do not suffer the same outcomes disadvantage.
Rural health care is a complex system with a variety of factors, but it seems that two lessons keep hitting us over the head: distance matters; and small, local services are better than no services.
Have a different take? Give this article a read and let us know what you think.
†Haraldsdottir S, Gudmundsson S, Bjarnadottir RI, Lund SH, Valdimarsdottir UA. Maternal geographic residence, local health service supply and birth outcomes. Acta Obstet Gynecol Scand 2015; 94: 156–164
‡Grzybowski S, Stoll K, Kornelsen J. Distance Matters: A Population Based Study Examining Access to Maternity Services for Rural Women. BMC Health Sci Res 2011; 11(147): 1-8