Autonomous artificial intelligence increases real-world specialist clinic productivity in a cluster-randomized trial.

Autonomous artificial intelligence (AI) promises to increase healthcare productivity, but real-world evidence is lacking. We developed a clinic productivity model to generate testable hypotheses and study design for a preregistered cluster-randomized clinical trial, in which we tested the hypothesis that a previously validated US FDA-authorized AI for diabetic eye exams increases clinic productivity (number of completed care encounters per hour per specialist physician) among patients with diabetes. Here we report that 105 clinic days are cluster randomized to either intervention (using AI diagnosis; 51 days; 494 patients) or control (not using AI diagnosis; 54 days; 499 patients). The prespecified primary endpoint is met: AI leads to 40% higher productivity (1.59 encounters/hour, 95% confidence interval [CI]: 1.37-1.80) than control (1.14 encounters/hour, 95% CI: 1.02-1.25), p < 0.00; the secondary endpoint (productivity in all patients) is also met. Autonomous AI increases healthcare system productivity, which could potentially increase access and reduce health disparities. ClinicalTrials.gov NCT05182580.

Managing two-dose COVID-19 vaccine rollouts with limited supply: Operations strategies for distributing time-sensitive resources.

Distributing scarce resources such as COVID-19 vaccines is often a highly time-sensitive and mission-critical operation. Our research was prompted by a significant obstacle that the United States and other nations encountered during the early months of the COVID-19 vaccination campaign: Most COVID-19 vaccines require two doses given 3 or 4 weeks apart. Given the severely limited supply and mounting pressure on many countries to reduce hospitalizations and mortality, how to effectively roll out two-dose vaccines was a critical policy decision. In this paper, we first model and analyze inventory dynamics of the rollout process under three rollout strategies: (1) holding back second doses, (2) releasing second doses, and (3) stretching the lead time between doses. Then we develop an SEIR (susceptible, exposed, infectious, recovered) model that incorporates COVID-19 asymptomatic and symptomatic infections to evaluate these strategies in terms of infections, hospitalizations, and mortality. Among our findings, we show releasing second doses reduces infections but creates uneven vaccination patterns. In addition, to ensure second doses are given on time without holding back inventory, strictly less than half of the supply can be allocated to first-dose appointments. Stretching the between-dose lead time flattens the infection curve and reduces both hospitalizations and mortality compared with the strategy of releasing second doses. We also consider an alternative single-dose vaccine with lower efficacy and show that the vaccine can be more effective than its two-dose counterparts in reducing infections and mortality. We conduct extensive sensitivity analyses related to age composition, risk-based prioritization, supply disruptions, and disease transmissibility. Our paper provides important implications for policymakers to develop effective vaccine rollout strategies in developed and developing countries alike. More broadly, our paper sheds light on how to develop effective operations strategies for distributing time-sensitive resources in times of crisis.

NIH funding of COVID-19 research in 2020: a cross-sectional study.

OBJECTIVE: This study aims to characterise and evaluate the National Institutes of Health's (NIH's) grant allocation speed and pattern of COVID-19 research. DESIGN: Cross-sectional study. SETTING: COVID-19 NIH RePORTER Dataset was used to identify COVID-19 relevant grants. PARTICIPANTS: 1108 grants allocated to COVID-19 research. MAIN OUTCOMES AND MEASURES: The primary outcome was to determine the number of grants and funding amount the NIH allocated for COVID-19 by research type and clinical/scientific area. The secondary outcome was to calculate the time from the funding opportunity announcement to the award notice date. RESULTS: The NIH awarded a total of 56 169 grants in 2020, of which 2.0% (n=1108) wwas allocated for COVID-19 research. The NIH had a US$45.3 billion budget that year, of which 4.9% (US$2.2 billion) was allocated to COVID-19 research. The most common clinical/scientific areas were social determinants of health (n=278, 8.5% of COVID-19 funding), immunology (n=211, 25.8%) and pharmaceutical interventions research (n=208, 47.6%). There were 104 grants studying COVID-19 non-pharmaceutical interventions, of which 2 grants studied the efficacy of face masks and 6 studied the efficacy of social distancing. Of the 83 COVID-19 funded grants on transmission, 5 were awarded to study airborne transmission of COVID-19 and 2 grants on transmission of COVID-19 in schools. The average time from the funding opportunity announcement to the award notice date was 151 days (SD: ±57.9). CONCLUSION: In the first year of the pandemic, the NIH diverted a small fraction of its budget to COVID-19 research. Future health emergencies will require research funding to pivot in a timely fashion and funding levels to be proportional to the anticipated burden of disease in the population.

The Association Between Timing of Elective Surgery Scheduling and Operating Theater Utilization: A Cross-Sectional Retrospective Study.

BACKGROUND: Overutilization of operating theaters (OTs) occurs when actual surgery duration exceeds scheduled duration, which could potentially result in delays or cancelations in subsequent surgeries. We investigate the association between the timing of elective surgery scheduling and OT overutilization. METHODS: A cross-sectional retrospective study was conducted using electronic health record data of 27,423 elective surgeries from July 1, 2016, to July 31, 2018, at a mid-Atlantic academic medical center with 56 OTs. The scheduling precision of each surgery is measured using the ratio of the actual (A) over the scheduled or forecast (F) length of surgery to derive the predictor variable of A/F (actual-to-forecast ratio [AF]). Student t test and χ2 tests analyzed differences between OTs reserved within and over 7 days of surgery for continuous and dichotomous variables, respectively. Hierarchical regression models, controlling for potential confounds from the hospital environment, clinicians' work experience and workloads, patient factors, scheduled OT length, and operational and team factors isolated the association between OTs reserved within 7 days of the elective surgery with AF. RESULTS: The Student t test indicates that OTs reserved within 7 days of surgery had significantly higher AF (1.13 ± 0.53 vs 1.08 ± 0.41; P < .001). In-depth Student t test analyses for 4 patient groups, namely, outpatient, extended recovery, admission after surgery, and inpatient, indicate that AF was only significantly different for OTs reserved within 7 days for the admission after surgery group (1.15 ± 0.47 vs 1.09 ± 0.35; P < .001) but did not reach statistical significance among the outpatient, extended recovery, and inpatient groups. After controlling for potential confounds, hierarchical regression for the admission after surgery group reveals that OTs reserved within 7 days took 2.7% longer than the scheduled length of surgery (AFbeta, 0.027; 95% CI, 0.003-0.051; P = .027). CONCLUSIONS: Elective surgeries scheduled within 7 days of surgery were associated with significantly higher likelihood of OT overutilization for surgical patients who will be admitted after surgery. Further studies at other hospitals and a longer period of time are needed to ascertain a potential "squeeze-in" effect.

Transforming COVID-19 vaccines into vaccination : Challenges and opportunities for management scientists.

Amid the prolonged COVID-19 pandemic, the miraculous breakthroughs of multiple effective and safe COVID-19 vaccines offer hopeful prospects. Yet, the endgame of the pandemic is not vaccines; it is vaccination. The daunting challenge of vaccinating the world offers ample investigative opportunities for management scientists who are interested in improving the efficiency and equity of vaccine supply chains. In this article, we provide a brief overview of these opportunities through three constituent parts: (1) supply, (2) demand, and (3) matching supply with demand.

CRC COVID: Colorectal cancer services during COVID-19 pandemic. Study protocol for service evaluation.

INTRODUCTION: COVID-19 has had an impact on the provision of colorectal cancer care. The aim of the CRC COVID study is to describe the changes in colorectal cancer services in the UK and USA in response to the pandemic and to understand the long-term impact. METHODS AND ANALYSIS: This study comprises 4 phases. Phase 1 is a survey of colorectal units that aims to evaluate adherences and deviations from the best practice guidelines during the COVID-19 pandemic. Phase 2 is a monthly prospective data collection of service provision that aims to determine the impact of the service modifications on the long-term cancer specific outcomes compared to the national standards. Phase 3 aims to predict costs attributable to the modifications of the CRC services and additional resources required to treat patients whose treatment has been affected by the pandemic. Phase 4 aims to compare the impact of COVID-19 on the NHS and USA model of healthcare in terms of service provision and cost, and to propose a standardised model of delivering colorectal cancer services for future outbreaks. ETHICS AND DISSEMINATION: This study is a service evaluation and does not require HRA Approval or Ethical Approval in the UK. Local service evaluation registration is required for each participating centre. In the USA, Ethical Approval was granted by the Research and Development Committee. The results of this study will be disseminated to stakeholders, submitted for peer review publications, conference presentations and circulated via social media. REGISTRATION DETAILS: Nil.