The following is an Afterword to a special issue on the Anthropology of Compliance, edited by Kalman Applbaum and Michael Oldani, in Anthropology & Medicine, Vol. 17, No. 2, August 2010, 245–247. Check it out!
This collection on new anthropologies of compliance edited by Kalman Applbaum and Michael Oldani illustrates the ongoing importance of careful, reflective qualitative research in contemporary biomedicine at every level. Compliance – itself a bureaucratic analytic for assessing health outcome trends – is studied from an impressive array of directions and methods. Nurses, doctors, adolescent and senior patients, NGOs, parents, and college student consumers are researched via ethnography, surveys, interviews, rhetorical analysis, and narrative techniques; and the illnesses considered include cancer, TB, depression, and diabetes. Out of this wide grid emerges a remarkably consistent set of insights, or omens, of a compliance crisis that is most tragic because it almost unregisterable.
Compliance’s usefulness as a measure of prescribed treatment adherence by a patient has been its ability to index treatment success, and therefore help in ensuring and governing health. What these anthropologists have discovered, however, is that because of its function as an index, compliance has been itself instrumentalized and made to serve goals sometimes quite far from health. Most fascinating is that it is not only pharmaceutical and insurance companies – the standard bugbears of capitalized medicine – that are engaging in this instrumentalization of compliance, but it is happening across the board, by the severely mentally ill, TB treaters, doctors and nurses.
In a surprising fashion, this collection of articles verifies Goodhart’s Law, named after the economist whose finding was that ‘Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.’ In other words, when rewards or punishments target an index rather than the outcome, actions quickly migrate toward maximizing the index independent of outcomes. Marilyn Strathern pointed anthropologists to this phenomenon in her work on audit culture, as it grossly affects our own academic scholarship as we are asked to meet performance indicators that quickly take on a life of their own. As Strathern (2000) put it, ‘It is very simple: you turn the system of measurement into a device that also sets the ideal levels of attainment. In short, audit measures become targets.’ The most frightening example of this in this collection is Oldani’s discovery that rewarding doctors directly for their compliance numbers almost immediately perverts their attention away from best patient care toward the relative value of that patient’s compliance score. This renders some patients not worth treating and others worth mislabeling. The flexible autonomy granted doctors in making decisions is in turn seen by the insurance company as a target of audit, oversight, and ultimately reduction. Thus, one clear corollary of compliance orientation is structural distrust of all actors.
Harper’s essay on TB compliance in the wake of multiple drug resistance finds that even in the face of rigorous studies showing that families are able to co-monitor each other as effectively as directly-observed treatment (DOT) protocols, it seems the lack of visible marked compliance cannot be trusted by those embedded within compliance measurement systems. Again, Strathern (2000) helps us see the processural dilemma: ‘The mechanism is already there in the focus on outcome: by bypassing descriptive observation, or rather by restricting the output (results) of observation to data suitable for constructing indicators.’ Compliance can therefore come to stand in as the only measure of effectiveness.
The second aspect of the instrumentalization of compliance charted by these anthropologists is the reinforcement of frames – discourses or common sense – such as the biopsychiatric model of mental illness as disease, the necessity of drugs for life, good drugs versus bad drugs, and noncompliance as the cause of poor health outcomes. Compliance as a fully operationalized term, concept and practice, fixes and reifies the assumptions on which it depends. Thus, Rouse found that doctors and patients were both unable to escape the logic within which non-compliance could lead to anything but poor health outcomes. This prevented any meaningful dialogue between them about possibly different views of health or the relative value of different treatments, especially about different economic conditions. Similarly, the operative axiom that going off of an antipsychotic must lead to the re-emergence of illness symptoms can effectively mask (or serve as an alibi for) the real possibility that a drug has withdrawal effects. Biopsychiatric frames – that mental illness is both verified and controlled only through pharmaceuticals – are found in both narratives of adolescent sufferers (Longhofer and Floersch) and in media accounts of murderers supposed to be suffering from mental illness (Glick and Applbaum). In both cases, medication compliance is treated as the linchpin that keeps a bad past and bad future at bay.
Where compliance may have been at times a research question of efficacy or an index correlating with health, throughout these ethnographies compliance has become the operative definition of risk prevention. At the experiential level, we might call this process ‘interpillation’. With a tragic nod to Althusser’s own struggles with mental illness, interpillation is the process of calling into being biomedical subjects as having been always-already in need of treatment. Individuals, especially those with mental illness, are increasingly left with no other subject position. Thus, Glick and Appelbaum quote a sufferer saying, ‘I could be a threat if I weren’t taking my medication.’ Indeed these ethnographies collectively suggest that health system(s) broadly speaking hail all of us into an ethics that starts with, rather than ends with compliance: where ‘failure to treat is an ethical dilemma’ (Longhofer and Floersch); where ‘many times, trusting is not a choice, patients can be forced to trust,’ (Liebing); where everyone – patient, doctor, and institution – becomes accountable to compliance (Harper; Oldani); where biopsychiatric viewpoints are strategically but also necessarily used (Brodwin; McKinney and Greenfield; Glick and Applbaum).
Overall, these findings are important, bleak, and tragic. These papers nonetheless offer possible solutions, neither easy nor unfortunately efficient. They actively ask us to reframe our sense that compliance is a measure. Liebing says it best: ‘Adherence is a matter of concern.’ Compliance, in other words should be the problem, not the solution. Compliance is (or should be) the busy intersection of illness variability, of worries, loneliness, caring, and trust. Compliance is the outcome of sufferers and providers put into terrible situations where treatment is not perfect, and where medications, environments, family members, infrastructures, drugs are also not perfect. Adherence is the call to attention, to attend to the ongoing problem, and the people (all the people) involved in it. What McKinney and Greenfield discovered among privileged college student consumers of antidepressants was the deep roles – symbolic, material-semiotic, family-systems, experimental-experiential – that com- pliance through medications, diagnosis, and side effects played in their lives. This same insight into deep roles runs through the findings regarding people with severe mental illness, fatal illnesses, tuberculosis, and other illnesses.
The policy we sorely need is therefore a constant revaluation of endpoints not indices. This means according to many of the authors, paying more attention to variability in patients, diseases, illness courses, medication responses, and quality of life desires. The horrifying blinders and straitjackets that instrumentalized compliance puts on everyone involved in compliance (think especially of Oldani’s doctor) produces an anti-medicine, and possibly even more illness. If you agree in principle with this, but still insist that practicality in mass health requires management by numbers, then attend to these papers as proof that Goodhart’s Law holds true in health as well. The numbers we would like (for good reasons) to manage by can quickly turn against us. We owe it to health and objectivity to have more than numbers.
Strathern, Marilyn. 2000. Abstraction and decontextualisation: An anthropological comment or: e for ethnography. At Virtual Society? Get Real! http://virtualsociety.sbs.ox.ac.uk/ GRpapers/strathern.htm
Contents of the Special Issue
Kalman Applbaum, Michael Oldani
Jeffrey Longhofer, Jerry Floersch
Kelly A. McKinney, Brian G. Greenfield
Patient and practitioner noncompliance: rationing, therapeutic uncertainty, and the missing conversation
Assessing the ‘relative value’ of diabetic patients treated through an incentivized, corporate compliance model
Michael J. Oldani
Douglas Glick, Kalman Applbaum