Regulatory innovation and (de)legitimisation: accountability challenges in the healthcare field

ABSTRACT

Accounting research examines accountability in the healthcare field but little attention is paid to how contemporary regulatory innovations (e.g. AI-driven data algorithms) impact accountability objectives. We address this concern through a study of how regulatory innovations may appropriate accountability by (de)legitimising healthcare actions. This is examined in the case of a regulatory innovation enacted in the US healthcare sector. Specifically, we employ a field study of a healthcare provider subject to oversight to highlight how actors (de)legitimise the provision of healthcare services. Our findings show that regulatory innovation generates debates around illegitimate services as determined by healthcare oversight contractors. These contractors are incentivised to appropriate AI-driven data algorithms towards a focus on reducing costs of non-standard healthcare services while healthcare providers legitimate those services as quality healthcare. Additionally, findings suggest that more extensive implementation of AI-driven data algorithms in the healthcare field may serve to limit healthcare options, particularly when combined with private contracting. This study contributes to the literature by showing the (de)legitimising consequences that regulatory innovations combining the use of algorithmic decision-making and private incentive contracts have for accountability and delivery of services in the healthcare field.

KEYWORDS:

• Regulatory technologies
• legitimisation
• incentive contracts
• government contracting
• healthcare
• accountability

Acknowledgements

the authors would like to thank the guest editors and anonymous reviewers, Yves Gendron, Ethan Lamothe, Thomas McKee, participants at the 2018 Public Interest midyear meeting, the AAA 2018 annual meeting, and the AAA 2022 annual meeting for their helpful insights and suggestions on earlier drafts of this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Proprietary information came from Freedom of Information Act (FOIA) requests. Per the FOIA, any person in the U.S. has the right to request access to federal government records or information unless protected by exemption.

2 We acknowledge that administrative managers might view scaling back operations (via reduced costs) as a success, while clinicians might view offering less services as a detriment (Malmmose & Kure, 2021; Rautiainen et al., 2022).

3 For instance, increased cost-constraints led the healthcare industry to seek out less costly treatment options, e.g., treating patients at home, giving rise to growth of the home health industry (Friedman, 2006).

4 For instance, initiatives focusing on financial outcomes can cause providers to lack compassion in treatment by viewing patients as cost objects to be processed and classified into “the profitable sick and the unprofitable sick” (Llewellyn et al., 2020, p. 19). From an organisational standpoint importing financial logics is sometimes referred to as mission drift, or moving away from initial accountability objectives (environmental, institutional, social, stewardship, etc.) to focus on financial goals such as profit maximisation (Ebrahim et al., 2014).

5 Both increased focus on financial goals, efficiency and effectives as well as outsourcing of functions that improve financial goals, efficiency and effectiveness are reminiscent of the philosophy of new public management where public sector organisations mimic systems used in the private sector (Modell, 2018).

6 We received appropriate institutional review board approval before conducting the field study.

7 Healthcare fraud estimates are as high as 10% of total expenses (Mashaw & Marmor, 1994), costing American taxpayers up to $3 trillion annually (Federal Bureau of Investigation, 2014). See also the Annual Reports titled “Healthcare Fraud and Abuse Control Program” published since 1997 by the Office of Inspector General (OIG) in conjunction with DHHS and the Department of Justice (DOJ).

8 Prior research states the importance of researchers acknowledging their “position in the field” (Pratt, 2009). None of the authors had a pre-existing relationship with any of the employees of the organisation in this study.

9 Public data shows that the provider operates in a region where the poverty level is approximately 40% (nearly three times the national average), median household income is less than US$20,000, and the unemployment rate is above the national average.

10 We italicise and label all interview quotes with the participant ID and their position according to Table 1, Panel C. We label data gleaned from our ethnographic observations or impromptu discussions only with the participant ID. These are not italicised.

11 Pre-payment review requires that claims (in this case 100% of claims) be reviewed before reimbursement can be made, extending the time and effort for claims processing and lengthening the provider’s cash flow cycle.

12 Section 1.6.1.1.2 includes the following language: CMS will evaluate ZPIC performance measures including, but not limited to: Completeness and accuracy of data analysis; Completeness and accuracy of BI investigations and cases; Completeness and accuracy of all deliverables; Completeness and accuracy of medical reviews; Completeness and accuracy of claims reviews; Completeness and accuracy of audits.

13 Traditional auditor incentives are designed around avoiding conflicts of interest and prohibiting contingent fees due to the risk of detrimental incentives, which could create goal incongruence.

14 We examined the percentage breakdown of the three compensation components presented in contracts and contract modifications received through a FOIA request. Compensation data for the contractor responsible for the zone where the healthcare provider in this study is located is comparable to compensation data for contractors responsible for other zones.

15 Reporting on implementation of the regulatory technology and contractor operations during the period of study are portrayed as successful based on an increasing return on investment of 3.3:1 in 2012; 5:1 in 2013; 10:1 in 2014; and 11.5:1 in 2015 (CMS, 2016; DHHS, 2012, 2014, 2015). Congress used these reports to justify expansion to other government programs, but since then has stopped issuing the reports.