Abstract
In Metrics: What Counts in Global Health, Vicanne Adams uses an anthropological perspective to present a collection of ethnographic explorations into the shift towards postcolonial perspectives on global health. She also interrogates an accompanying trend towards certain types of quantitative metrics, based on statistical measures, experimental models and cost-effectiveness perspectives. She argues that ‘global health metrics pose a problem of knowledge in relation to our understanding … of global health efficacy’ (p. 6) and criticizes the trend towards randomized controlled trials (RCTs) as ‘arbiters of what counts in relation to health outcomes’ (p. 6). Questions are raised as to whether the emphasis on metrics, in fact, constitutes a political position by claiming to be ‘neutral’.
Adams explores the 19th century development of global ‘universal standards’ to measure health, presenting this as a colonial project that extended into the 20th century when international agencies such as WHO became established. This phase focused on emphasizing econometric measurements and presenting less advantaged parts of the globe as ‘vulnerable’ and dependent. She argues that metrics are presented as a panacea to the political obstacles to achieving better ‘health for all’ at the global scale and that schemes such as the Millennium Development Goals reinforced measurement using indicators such as disability adjusted life years. There is criticism of what is represented as an overemphasis on cost-effectiveness, ‘scalability’ and the risk of focusing solely on interventions which are amenable to assessment through randomized controlled trials. Speaking of metrics as the ‘sovereign forms that (directly and indirectly) aspire to govern in our times’ (p. 49), Adams argues that alternatives, such as more emotive ‘stories’, may help inspire us to ‘rethink’ metrics.
Contributions to this collection by Claire Wendland and Adeola Oni-Orisan examine the substantive issue of measurement of maternal mortality in African states, in ways which may not represent reality on the ground and may be influenced by political considerations. Wendland emphasizes the reliance on estimates rather than ‘hard statistics’ and that the precise interpretation of some definitions of key parameters leaves them open to criticism as potentially inaccurate measures of the aspect of health outcomes or health care that they are designed to assess. Also highlighted are difficulties of summarizing, at the level of global regions or for all countries, when in some cases data are missing over long periods. She highlights the failure to meaningfully assess quality of care or quality of life using metrics. Referring to ‘entanglements of politics and metrics’ (p. 85), Oni-Orisan discusses what seemed to be reluctance on the part of some health care institutions to record deaths among patients referred to them. The author criticizes the pressure on NGOs providing health care to present positive statistical reports that correspond to the requirements of major governmental and private sector donors, and financial incentives offered to health care providers, in some cases, to meet certain quantified health care goals.
Marlee Tichenor and Molly Halles focus on ‘Metrics Politics’. Their arguments suggest that the production of sound statistical data may be ‘hostage’ to political processes locally and internationally. Tichenor points to the literature contrasting global health security considerations with concerns about humanitarian biomedicine. The dependence on health care staff to provide statistics is presented as raising issues of conflicting interests, accountability and citizenship. In localities where health needs are most pressing, useful and meaningful data are most likely to be lacking. Halles examines the situation for native American populations. It is argued that measures intended to address problems of access to health care should not be confused with metrics indicating health outcomes of interventions. It is argued (p. 137) that ‘The use of health metrics may undermine … efforts to affirm the legitimacy of local knowledge’.
Chapters by Susan Erikson and Lilly Walkover suggest that neoliberal, private sector structures may influence the growing emphasis on quantifiable evidence and on push for aspects of health which can be quantified and evaluated by economic criteria such as cost-effectiveness. Erikson stresses the problem of assessing ‘value’ in complex systems such as health care provision. As in other chapters, specific formulae are presented and each component of each formula is critiqued in terms of whether it is meaningful. It is pointed out that injection of funds may not always provide a way to improve health outcomes, and that health inequality is not well addressed by market-based approaches. Walkover draws attention to the difficulties of attributing health outcomes to particular interventions using statistical methods, and the inability of statistics to communicate individual experience and knowledge.
Carolyn Smith-Morris and Pierre Minn discuss how metrics can be construed as a particular type of ‘story’, involving fidelity to protocols or presentation of health and health care, that appeals to a particular type of agency and funder. Smith-Morris comments on the contrasting impressions of the same interventions gained from anthropological ethnographic reports or from RCTs structured by CONSORT guidelines, which follow strict conventions regarding the design of randomized trials. Minn points out that accountability methods may discourage ‘non-standard’ interventions, although these may be beneficial when extreme or unusual conditions impact on health.
Adams concludes with a discussion of novel approaches which challenge what is presented as the ‘tyranny’ of metrics (p. 17), arguing that ‘Metrics enable certain kinds of medical practices while impeding others’ (p. 225).
This book provokes a thoughtful debate concerning the sociopolitical cultures promoting perspectives based on metrics. However, it is perhaps unfortunate that in doing so, it does not acknowledge that the more ethnographic approach advocated here also can be ‘disproportionately’ powerful (for example, specific events and reports of experiences of individuals often can also prove influential upon sociopolitical actions). Ethnographic reports have their own limitations in guiding action, since they do not give a sense of how generally distributed aspects of population health may be. Also, it might be argued that some statistical methods using Bayesian and hierarchical modelling techniques are more sensitive to varying scenarios and local conditions than the authors of this book seem willing to acknowledge. More emphasis on the complementarity of quantitative and qualitative methods might have provided a more balanced account.