Research

Select Publications

Singh, M. (2021). Heuristics in the delivery room. Science, 374(6565), 324-329.

Abstract

Clinical decisions made in the delivery setting are often made under high pressure, great uncertainty, and have serious consequences for mother and baby. Theories of decision-making suggest that individuals in such settings may resort to using “heuristics”, or simplified decision-rules, to aid complex decision-making. This study investigates whether physicians’ delivery-mode decisions (i.e., when to perform a vaginal vs. a cesarean) are influenced by such a heuristic. Electronic health record data spanning 86,000 deliveries suggests that, if the prior patient had complications in one delivery-mode, the physician will be more likely to switch to the other -- and likely inappropriate -- delivery-mode on the subsequent patient, regardless of patient indication. There is evidence that this heuristic has small, suboptimal effects on patient health.

Results as a limerick



A patient has a bad reaction,

To the Doc’s deep dissatisfaction,

Feeling unfit,

Doc cries out, "Oh sh*t!"

And switches her medical plan of action.

Press: New York Times Upshot, LA Times, The Conversation, Quartz, Reuters Health, Technology Networks, Yahoo Finance, MedicalXpress, American Council on Science and Health, NewsWise, PressReleasePoint, The Deccan Her- ald, DevDiscourse, Haaretz, Arizona Daily Star, Dispatch Argus, The Buffalo News, Lincoln Journal Star, Bozeman Daily Chronicle, Yahoo News, Fairfield Citizen, Seattle Post-Intelligencer, Medical Xpress, Pourquoi Docteur, DNYUZ, Mirage News, The Medical News, KRQE, Latestly, Foreign Affairs New Zealand, St. Louis Post-Dispatch, GoSkagit, Entrepreneur, Bismarck Tribune, KULR, The Darien Times, Shelton Herald, New Canaan Advertiser, San Antonio Express News, Idaho Press, La Vanguardia

Click HERE for an ungated link to the paper

Working Papers

Capacity Strain and Racial Disparities in Hospital Mortality (with Atheendar Venkataramani)

NBER WP 30380 (Revisions submitted to Nature Communications)

Abstract

A growing literature has documented racial disparities in health care. We argue that racial disparities may be magnified when hospitals operate at capacity, when behavioral and structural conditions associated with poor patient outcomes – e.g., limited provider cognitive bandwidth or reliance on biased care algorithms – are aggravated. Using detailed, time-stamped electronic health record data from two large hospitals, we document that in-hospital mortality increased more for Black patients than for White patients when hospitals approached capacity. We estimate that 8.5% of Black patient deaths were capacity-driven and thus avoidable. We then investigate the extent to which differential care inputs explain our findings. While strain exacerbated wait times similarly for Black and White patients, Black patients both waited the longest at high strain and faced greater mortality consequences from prolonged wait times. Finally, the largest racial disparities in mortality were among women and uninsured patients, highlighting biases in provider behavior and hospital processes as key mechanisms driving our results.

Results as a limerick




On clinical capacity does care depend

whether or not providers intend,

Hospitals get busy,

all systems a-tizzy,

with Black patients harmed most in the end.


Behavioral Responses to Surgeon Report Cards (with Jacob Zureich) (Revisions requested at Management Science)

Abstract

Feedback interventions such as report cards are often used in healthcare to encourage physicians to improve performance. However, providing feedback is tricky because it often conveys negative information to individuals about their performance, which can induce behavioral responses (e.g., dejection, confusion, attributional errors, etc) that interfere with learning and improvement. In this paper, we use novel data from a surgeon report card (linked to 320,000 surgeries) to examine surgeon response to positive vs. negative feedback, and to identify which surgeons improve most from such feedback. Exploiting two sources of plausibly exogenous variation in report card information, we highlight two results: i) consistent with behavioral responses to feedback, patient outcomes improve for surgeons receiving positive feedback and deteriorate for those receiving negative feedback, and ii) the largest returns to the feedback accrue to those already skilled at learning from experience ex-ante (i.e., the inframarginal physician) because they respond more effectively to both positive and negative feedback. These results are replicated in the lab, where we explore mechanisms and test interventions to attenuate suboptimal behavioral responses, such as providing help interpreting feedback and adding improvement-based incentives. Overall, report cards should be carefully designed to avoid triggering counterproductive behavioral responses to negative feedback.


Results as a limerick





Are report cards a fad?

Or could they help just a tad?

We find feedback is tricky,

‘cuz priors are sticky

And the good helps more than the bad.


Abstract


Power, defined as the asymmetric control of valued resources, affects most human interactions. Yet there is little observational evidence on how power affects real-world behavior and resource allocation. We examine this question using the power differential in the doctor-patient encounter: while it favors the physician in the clinical setting, powerful patients may be able to reduce this asymmetry and influence physician behavior. We exploit the quasi-exogenous assignment of 1.5 million patients to physicians in US military emergency departments, using their difference in military ranks to measure their power differential. We find that, keeping patient rank constant, “high-power” patients (i.e., who outrank their physician) receive greater effort and resources from physicians and have better outcomes than “low-power” patients. We also find that: (i) low-power patients suffer on days their physician cares for a high-power patient; (ii) doctor-patient concordance on race and sex is important to such power dynamics; and (iii) within-physician effort is higher for patients recently promoted than those about to be promoted. Thus, power confers nontrivial advantage to its possessor. This is particularly concerning in the healthcare setting, where such variation in care leads to inefficiencies and harms are borne by society’s most vulnerable patients.


Results as a limerick





Is power too beckoning a call?

Since doctors are human after all,

We took the military’s example

Rank predicted care in our sample,

The mighty were heeded o’er the small.



Strained and Constrained: How ICU Capacity Affects Physician Decisions (with David Howard and Thomas Valley)

Abstract

Limited ICU capacity has been said to exacerbate several consequences of COVID-19, such as mortality, misallocation of resources by physicians, and harmful spillovers on nonCOVID patients. However, whether greater ICU capacity would have avoided these pitfalls is not clear. The goal of this paper is to shed light on how ICU capacity affects physician decision-making and patient welfare when the healthcare system is strained. We use two sources of variation in ICU capacity to estimate its causal effect on physician thresholds for admitting patients to the ICU: i) ICU “expansions”, resulting from increases in hospital number of ICU beds, and ii) ICU “strain”, resulting from random fluctuations in ICU bed availability. Our analysis uses 100% inpatient EHR data from two hospitals (150K encounters. 2015-18) both of which expanded their ICUs separately. Importantly, we use lab test results to assign patients an objective and validated measure of ICU-need, called the eSOFA score. We motivate our empirical analyses using a model of physicians’ dynamic admission decisions when there is uncertainty about a patient’s ICU-need. Results show that increases in capacity cause physicians to lower their thresholds for ICU admission, with mixed effects on patient welfare. Patients are 1.8 pp more likely to be admitted to the ICU post-expansion (even when ICU strain is at pre-expansion levels), but there is significant heterogeneity in likelihood of admission and in-hospital mortality by ICU strain and ICU-need. Expanding ICU capacity does not always help the patients who need it most, and at times harms patients who need it least. Finally, the greatest benefits of ICU expansions accrue to patients in the general wards by allowing physicians to better allocate non-ICU resources, especially when the hospital is operating at capacity.