D ATA D R I V E N . P O L I C Y F O C U S E D LDI ResearchBRIEF Research to Improve the Nation’s Health System 2020 . No. 2 ARE THERE ENOUGH EXPERIENCED PHYSICIANS TO TREAT PATIENTS HOSPITALIZED WITH COVID? Kira L. Ryskina, MD, MSHP Journal of General Internal Medicine – published online September 19, 2020. Healthcare – published online October 22, 2020. KEYFINDINGS In this national study of 438,895 physicians, 45% provided care to hospitalized patients and 7% provided critical care. At the high estimate of patients requiring hospitalization at the projected peak of the pandemic, 18 states and Washington, DC would have patient to physician ratios greater than 15:1 (a level associated with poor outcomes among hospitalized patients). There was considerable geographic variation in the availability of physicians: 41% of hospital service areas did not have a physician with critical care experience. THE QUESTION service areas in the U.S. The authors considered critical care experience, minimum visit volume, as well as risk factors for severe COVID (e.g., With the third wave of COVID threatening the U.S., hospitals are age greater than 65). In another, they calculated patient volume per anticipating another surge of patients. During the earlier waves, hospitals physician at the peak of the pandemic for each state and Washington, and health systems increased physician staffing by deploying available DC using projections from the Institute of Health Metrics and Evaluation physicians to attend on the inpatient wards. However, little is known about model. The estimates accounted for other factors that affect the ratios, the experience of physicians with treating hospitalized patients, including such as transmission rates among health care workers and reduction in those who require critical care. In fact, over the past decade hospital hospitalizations for non-COVID causes. care became increasingly concentrated among physicians who practice there exclusively (i.e., hospitalists). How many US physicians have recent experience treating hospitalized patients, and how well does physician THE FINDINGS supply match regional projections of need during future COVID peaks? Fewer than half of physicians in the study billed Medicare for hospital Using 2017 Medicare data, Ryskina and colleagues measured the number visits in 2017. One in five physicians billed for 50 or fewer hospital visits of physicians who billed Medicare for services provided to hospitalized in a year. Of the physicians with recent experience treating hospitalized patients, the number of physicians who provided critical care, and the patients, one in ten was over 65 years of age. number of physicians regardless of experience. The authors completed If only physicians with experience treating hospitalized patients are two separate analyses. In one, they measured the number of physicians utilized, the median patient volumes across all states are 10 to 13 patients with relevant experience per 100,000 residents in the 3,360 hospital per physician. Eighteen states and Washington, DC would have patient COLONIAL PENN CENTER | 3641 LOCUST WALK | PHILADELPHIA, PA 19104-6218 | LDI.UPENN.EDU | P: 215-898-5611 | F: 215-898-0229 | @PENNLDI ResearchBRIEF LDI Ryskina et al .: US Physicians Managing Hospitalized Older Adults Table 1 Characteristics of Physicians Who Treated Hospitalized Patients Overall and Those Who Provided Critica Physicians who treated hospitalized Physicians who Figure 1. Projected Patient Volume per Physician at the Peak of the Pandemic: to physician ratios above patients (N =15:1 at the projected peak of the 196,788) care patients ( Medical and Surgical Specialties Characteristics* pandemic (Figure n 1). If all physicians were % deployed regardlessn of Gender experience, each provider would need to treat a median of 5 to Male 6 patients, with 147,496 only one state – Georgia75.0 – exceeding 15 patients24,429 Female 49,292 25.0 7218 per provider. Years from medical school graduation† There was considerable geographic variation in the 0–10 11–20 availability of physicians: 19,675 63,817 41% of hospital 32.8 service areas did not have 10.1 3202 12,162 21–30 31–40 a physician with50,267 critical care experience (Figure 40,957 25.8 21.0 2). 8161 5542 41 + 19,902 10.2 2051 By volume of visits‡ Ryskina et al .: US Physicians Managing Hospitalized Older Adults 11–50 visitsJGIM 51–250 visits THE IMPLICATIONS34,132 66,392 17.3 33.7 13,669 13,407 250–500 visits 35,425 18.0 3360 Table 1 Characteristics of Physicians Who Treated Hospitalized Patients Overall and Those Who Provided Critical Care in 2017 501–1000 visits These studies identify 32,000 potential shortages in physicians with recent 16.3 1058 Physicians who treated hospitalized patients (N = 196,788) Physicians who 1001 care patientsTop treated experience treating hospitalized or critically ill patients, especially in + critical specialties§ (N =531,647) 28,839 14.7 153 Characteristics* n % n Internal the areas of the 51,871 medicine Family practice % country currently facing 26.4 17,054 a resurgence of COVID. 8.7 9822 2212 Gender Male 147,496 75.0 24,429 Cardiology Patient to physician 77.2 16,501ratios greater than 15:1 8.4 have been associated 2174 General surgery 12,884 6.7 1708 Female Years from medical school graduation† 49,292 25.0 7218 with poor outcomes 22.8 Pulmonary/critical care 11,673in prior studies. Ensuring 5.9 sufficient physician 9484 supply may require widening the net to include all available 6247 ‡ 0–10 19,675 10.1 3202 Other 10.3 86,805 44.1 11–20 63,817 32.8 12,162 39.1 21–30 50,267 25.8 8161 26.2 31–40 41 + 40,957 19,902 21.0 10.2 5542 2051 *p value < 0.001 †Year of graduation 6.6 physicians 17.8for all comparisons acrossregardless categories ofof experience. physician Thus, health systems should characteristics from medical school was available for 194,618/196,788 physicians (1.1% missing) consider implementing training strategies to prepare physicians ‡ By volume of visits ‡“Other” category includes anesthesiology, medical and surgical subspecialists, medicine/pediatrics, orthopedic surgery, neur 11–50 visits 34,132 17.3 13,669 43.2 obstetrics and gynecology, otolaryngology, preventative medicine and rehabilitation, and urology not add up in advance of deployment. States should develop strategies to 51–250 visits 66,392 33.7 13,407 42.4 § 250–500 visits 35,425 18.0 3360 Percentages do10.6 to 100 due to rounding address regional shortages, particularly in critical care, such as 501–1000 visits 32,000 16.3 1058 3.3 1001 + 28,839 14.7 153 0.5 § Top 5 specialties Source: 2. A) low Internal Healthcare, estimates medicine 2020. of patients per medicine/surgical 51,871 provider with hospital 26.4 experience at the 9822peak of pandemic. B)31.0 High estimates telemedicine consultation of patients per medicine/ and additional resources for inter- ical provider with hospital experience at the 16,501 peak of pandemic. Family practice 17,054 8.7 2212 7.0 Cardiology General surgery 12,884 8.4 6.7 2174 1708 6.9 5.4 hospital transfers. Pulmonary/critical care 11,673 5.9 9484 30.0 sicians Other would be quarantined. We used 86,805 a conservative estimate44.1of not adjust the 6247ICU bed occupancy19.7 rate when estimating the number of ‡ sonable*pphysician value < 0.001 work schedulesacross to account for self-care and other non-COVID patients requiring critical care, because critically ill patients THE STUDY for all comparisons categories of physician characteristics †Year of graduation from medical school was available for 194,618/196,788 physicians (1.1% missing) sician ‡“Other” responsibilities - we category includes estimatedmedical anesthesiology, thatandeach physician surgical would subspecialists, are unlikely medicine/pediatrics, to avoid orthopedic surgery, hospitals. neurosurgery, To determine the number of COVID neurology, k everyFigure and gynecology, otolaryngology, preventative medicine and rehabilitation, and urologypatients in the hospital and ICU in each state at the peak of the § 4th shift. Thus, 25% of all physicians would be on service in Percentages 2. Physicians who Billed for Critical Care per 100,000 Hospital Service obstetrics do not add up to 100 due to rounding hospital during any given shift at the peak of the pandemic. The authors used the Medicare Provider Utilization and Payment pandemic, we used the Institute for Health Metrics and Evaluation AmericanArea Residents Hospital Association data was used to determine the (IHME) Covid-19 state projections (released onData to 2,identify October physicians in medical and surgical specialties who 2020), which mber of patients hospitalized and requiring ICU care at the peak of the assumed continued easing of current social distancing billed guidelines.7for acute hospital visits and critical care visits in Medicare demic.5 To do so, we multiplied the number of hospital and intensive e unit (ICU) beds available in each state by the average occupancy 2017 (the most recent data available). Physicians in specialties of those beds. We assumed a 38% reduction from historical hospi- 2.1. Analysis unlikely to be deployed as the primary physician for hospitalized zation rates to estimate the number of hospitalized non-COVID pa- patients during the pandemic were excluded: dermatology, nuclear ts, based on reported reduction in all-cause hospitalizations.6 We did IHME COVID projections were added to the estimates of the number medicine, of non-COVID patients to determine how many psychiatry, patients would be ophthalmology, pathology and radiology. Physicians trained in emergency medicine who are likely to be 3 needed in the emergency department were also excluded. One limitation of the study was that it did not consider trainees (e.g., residents) and advanced practitioners (e.g., nurse practitioners) who provide direct patient care to hospitalized patients. K.L. Ryskina, A. Bhatla, R.M. Werner. US Physicians with Recent Experience Managing Hospitalized Older Adults: An Observational Study. Journal of General Internal Medicine – published online September Figure. 1 Prevalence of physicians 19, 2020. who treated hospitalized Medicare beneficiaries requiring critical care per 100,000 ho Source: JGIM of physicians who treated hospitalized Medicare beneficiaries requiring critical care per 100,000residents Figure. 1 Prevalence in 2017. hospital service area The figure shows the geographic distribution of physicians who billed Medicare for critical care services residents in 2017. The figure shows the geographic distribution of physicians who billed Medicare for critical care services in 2017 per 100,000 residents in each hospital service area. A. Bhatla, K.L. Ryskina. Hospital residents in eachand ICUservice hospital Patient Volume Per area. Physician at Peak of COVID Pandemic: State-level Estimates. Healthcare – published online October 22, 2020. LEAD AUTHOR DR. KIRA L. RYSKINA Kira L. Ryskina, MD, MSHP is an Assistant Professor of Medicine in the Division of General Internal Medicine at the University of Pennsylvania. Her research aims to elucidate non-clinical factors that systematically influence physician practice of high-value care for older adults with multiple chronic conditions. Her recent work explores the role of physician specialization in the outcomes, costs, and care experience of patients receiving post-acute care in nursing homes. Dr. Ryskina completed clinical training in internal medicine and primary care at New York Presbyterian – Weill Cornell. She treats hospitalized patients at the Penn Presbyterian Medical Center in Philadelphia.