Data from 38,193 patients showed that managed care patients have COVID-19 risk factors similar to those of the general population and that a population health program decreased mortality.
Objectives: To determine (1) factors linked to hospitalizations among managed care patients (MCPs), (2) outcome improvement with use of outpatient off-label treatment, and (3) outcome comparison between MCPs and a mirror group.
Study Design: Retrospective cohort study comparing MCPs with an age- and gender-matched mirror group in Florida from April 1, 2020, to May 31, 2020.
Methods: A total of 38,193 MCPs in a Florida primary care group were monitored for COVID-19 incidence, hospitalization, and mortality. The highest-risk patients were managed by the medical group’s COVID-19 Task Force. As part of a population health program, the COVID-19 Task Force contacted patients, conducted medical encounters, and tracked data including comorbidities and medical outcomes. The MCPs enrolled in the medical group were compared with a mirror group from the state of Florida.
Results: The mean (SD) age among the MCPs was 67.9 (15.2) years, and 60% were female. Older age and hypertension were the most important factors in predicting COVID-19. Obesity, chronic kidney disease (CKD), and congestive heart failure (CHF) were linked to higher rates of hospitalizations. Patients prescribed off-label outpatient medications had 73% lower likelihood of hospitalization (P < .05). Compared with the mirror group, MCPs had 60% lower COVID-19 mortality (P < .05).
Conclusions: MCPs have risk factors similar to the general population for COVID-19 incidence and progression, including older age, hypertension, obesity, CHF, and CKD. Outpatient treatment with off-label medicines decreased hospitalizations. A comprehensive population health program decreased COVID-19 mortality.
Am J Manag Care. 2021;27(6):234-240. https://doi.org/10.37765/ajmc.2021.88595
The COVID-19 pandemic continues to affect populations across all sectors of society. A broad-based and multifaceted approach is necessary to manage at-risk populations. Multiple studies have shown that mortality significantly increases with age and comorbidities.1,2 A recent modeling study accounting for age, sex, and 11 underlying conditions, across 188 countries, estimates that 1.7 billion individuals globally have at least 1 underlying condition, putting them at increased risk of severe disease if infected with SARS-CoV-2, the virus that causes COVID-19.3 In particular, patients 60 years and older tend to have a more prolonged disease course, higher risk of complications, and significantly increased mortality.4 When combined with the sporadic access to care among the elderly, COVID-19 prompted the necessity for a coordinated care model, because by the time these adults seek care, they may already be severely ill due to both COVID-19 and any unmanaged chronic conditions.5
In one of the first evidence-based recommendations for individualized treatment for COVID-19, our group reviewed the literature and published a management approach based on disease course.6 However, there is still no outpatient treatment that is proven to reduce mortality. It has been established prior to COVID-19 that high-intensity primary care results in lower mortality rates.7-9 In these models, patients often perceive their doctors to be more vested in their health outcomes.10 This in turn motivates the patient to take on a more active role in their own health, resulting in personalized care tailored to the specifics of the patient.11 As an added benefit, medical cost reductions may also be achieved.12 However, little attention has been paid to outpatient management strategies with a population health–based approach to disease management.
To date, most COVID-19 clinical studies have focused on medical treatment, particularly in a hospital setting. Furthermore, it is unclear whether early intervention in an outpatient setting with an interdisciplinary, customizable approach that connects clinical practice, policy, and technology via a population health platform is beneficial. In this study, we conducted a retrospective review of a population health–based outpatient program undertaken from April 1 to May 31, 2020, in a mostly elderly and high-risk managed care population in Florida. We compared the results with those of an age- and gender-matched mirror population to test our hypothesis that managed care patients (MCPs) would fare better than the comparable population, primarily due to the model of care that proactively focuses on keeping patients healthy and not just treating disease.
Population Health–Based Patient Outreach and Outpatient Management
From April 1, 2020, to May 31, 2020 (study period), the health of 38,193 MCPs (insured by Medicare Advantage or managed Medicaid) under the care of Cano Health, a Florida-based primary care group (“medical group” [MG]) were tracked on a daily basis for COVID-19 incidence, hospitalization, treatments, and mortality. Beginning on April 6, 2020, all 38,193 patients were contacted by telephone to complete a COVID-19 survey. A total of 29,126 unique patients (76% of the study cohort) completed at least 1 survey during the study period. The survey protocol called for subsequent restratification surveys and rescoring every 2 weeks. Although 9067 (23.7%) patients did not complete the survey, they could have been referred to the COVID-19 Task Force (CTF) through an encounter with one of the MG’s providers (eAppendix Figure 1 [eAppendix available at ajmc.com]). Among patients who received a diagnosis of COVID-19, 96 (73%) completed the survey (eAppendix Table 1).
Based on the COVID-19 survey score, patients were stratified into 1 of 4 risk groups: low (0-10), medium (11-29), high (30-40), and very high (≥ 41) (eAppendix Table 2). Patients with a score of at least 30 or who were positive for COVID-19 were assigned to the CTF. The CTF was a team of clinicians and other health care professionals assigned to manage patients. The CTF utilized a COVID-19 treatment protocol based on the findings of our previous study.6 Each patient’s health status was continually monitored and updated, and treatment of underlying comorbidities was optimized.
Patient monitoring frequency was carried out based on their most recent risk score. Patients continued to be monitored until their symptoms resolved or a CTF provider discharged them from the CTF program (eAppendix Table 3). CTF protocols involved 3 general categories: (1) intensive patient monitoring (with frequent telephone encounters [“televisits”]), (2) medical optimization (such as optimization of treatment of chronic conditions and management of exacerbations), and (3) use of COVID-19 treatment protocol (eAppendix Table 4 and eAppendix Figure 2). These 3 interventions taken by the CTF are collectively referred to as the population health program (PHP).
The population was tracked for a number of clinical outcomes, including hospitalization due to COVID-19, hospital length of stay, intensive care unit (ICU) admission, COVID-19–specific mortality, all-cause mortality, and COVID-19 recovery. COVID-19 recovery was defined as a patient who, after being tracked for at least 14 days, was not hospitalized for and did not die of COVID-19. Poor outcomes, as a composite end point, were defined as hospitalization for COVID-19 for at least 14 days, ICU admission, or all-cause mortality.
A comparison analysis was performed of the cumulative incidence of hospitalizations, overall mortality, and inpatient mortality due to COVID-19 for patients managed by the MG. Our control group consisted of patients not affiliated with the MG who resided in the Florida counties where the MG operates: Miami-Dade, Broward, Palm Beach, Hillsborough, Osceola, and Orange. Data from Florida’s Department of Health were used to compile the age- and gender-matched mirror group.13
We examined the distribution and dispersion of data through descriptive numerical summaries and graphical tools such as scatter plots and box plots. In summary tables, the arithmetic mean and SD are presented. In summary tables of categorical variables, counts and percentages are used. For our bivariate analyses, we employed χ2 or Fisher’s exact tests to look at associations between study variables. For our multivariate models, we used generalized linear models to control for covariates such as age and gender. Stata (version 16.1), R (version 4.1.0), and MedCalc (version 19.2.3) statistical packages were used for descriptive calculations, group comparisons, and regression modeling. All statistical testing was carried out at the 5% (2-sided) significance level.
Data sources included hospital information exchange alerts, Florida databases, medical records, and the COVID-19 surveys. The population of MCPs studied is comparable with Florida’s Medicare and Medicaid population in terms of comorbidities. It should be noted that Florida’s Medicare population has higher rates of comorbidities than does the national population (eAppendix Table 5).14 Further, the characteristics of the Medicaid population in Florida are consistent with studied characteristics of the broader national Medicaid population.15
To explore this relationship between off-label outpatient treatment and clinical outcomes, 4 regression models were created. The first model used all-cause mortality as a dependent variable, with outpatient treatment, age, gender, and chronic kidney disease (CKD) as the independent variables. The second model used COVID-19 mortality as a dependent variable, with outpatient treatment, age, gender, and CKD again as the independent variables. The third model predicted hospitalization as a function of outpatient treatment, age, gender, and CKD. The final model predicted hospital length of stay as a function of outpatient treatment, age, gender, and CKD (eAppendix Table 6).
The statistical analysis was designed to test our research questions regarding the utility of a customized PHP during the COVID-19 pandemic: (1) What factors are linked to COVID-19 hospitalization among MCPs?; (2) Did outpatient treatment using FDA-approved off-label medications improve outcomes?; and (3) Did MCPs in the MG have better outcomes compared with an age- and gender-matched mirror group in the state of Florida?
Question 1: Factors Linked to COVID-19 and COVID-19 Hospitalization Among MCPs
Of the total tracked population (N = 38,193), 132 (0.35%) contracted SARS-CoV-2 and received a diagnosis of COVID-19. Data on age, gender, and comorbidities for the MCPs were obtained for 32,561 patients. Between those referred through the CTF surveys and those referred by the MG’s care management department, a total of 645 patients (1.7% of population) were referred to the CTF clinicians (eAppendix Figure 1).
Among patients with COVID-19, there were no significant differences in COVID-19 hospitalization or mortality between those who did and did not complete the survey. However, those who completed the survey were older, were more likely to be Hispanic, and had higher rates of chronic heart failure (CHF), yet had lower all-cause mortality (eAppendix Table 1). Irrespective of CTF referral, all patients in the study cohort were tracked for COVID-19 incidence, hospitalization, and mortality. All patients had access throughout the study period to telemedicine, prescription home delivery, and 24/7 care management.
The general MCPs and the COVID-19 subset were similar in most aspects; however, patients with COVID-19 were significantly older (mean age,70 vs 66 years; P < .001), were more likely to be Hispanic (87% vs 75%; P < .001), and were more likely to have hypertension (81% vs 49%; P < .001) (Table 1). Overall, among patients with COVID-19, 38% were hospitalized. The hospitalized patients with COVID-19 were significantly older (mean [SD] age, 73.6 [10.6] years) than nonhospitalized patients with COVID-19 (68.1 [10.4] years) (P < .01). We also found that a significantly greater proportion of hospitalized patients with COVID-19 had obesity (36% vs 17%; P < .05), hypertension (92% vs 74%; P < .05), CKD (52% vs 22%; P < .01), and CHF (32% vs 9.8%; P < .01) (Table 2).
Question 2: Off-label Outpatient Treatment Impact on COVID-19 Outcomes Within MG
Overall, no difference in age was found between the usual care (UC) (mean [SD] age, 72.1 [12.1] years) and off-label outpatient treatment (Tx) (69.2 [10.0] years) groups. UC is defined as outpatient nonspecific treatment that did not include off-label medications. A significantly greater percentage of UC patients were male (63% vs 38%; P < .05) and had CKD (54% vs 24%; P < .05); otherwise, no difference in demographics or comorbidities was found between the 2 groups (Table 3 [part A and part B]). No difference in hospital length of stay was discovered between the UC (mean [SD], 8.9 [11.0] days) and Tx (mean [SD], 12.4 [9.4] days) groups (eAppendix Figure 3).
Overall, the results show no difference in all-cause mortality, COVID-19 mortality, or hospital length of stay between the Tx and UC groups while controlling for age, gender, and CKD. However, we found a significant difference in hospitalization between the Tx and UC groups while controlling for age, gender, and CKD (n = 102; χ25 = 30.4; P < .001) (eAppendix Table 6). Specifically, Tx group patients were 0.27 (95% CI, 0.14-0.50) times less likely to be hospitalized than UC patients (eAppendix Figure 4).
Question 3: Differences Between MG Patients and Age- and Gender-Matched Controls in Florida
We compared the MG cumulative incidence and outcomes for COVID-19 with those of the state of Florida (eAppendix Tables 7 and 8). COVID-19 incidence and hospitalization rates were statistically equivalent (eAppendix Figures 5 and 6). Results from the multivariate logistic models reveal that after controlling for age and gender, MCPs in the MG were 0.40 (95% CI, 0.18-0.89) times less likely to die than control patients in Florida (n = 8531; χ23 = 495.55; P = .045). When we included hospitalization, we found that hospitalized MCPs were 0.37 (95% CI, 0.16-0.86) times less likely to die than control patients in Florida (n = 6765; χ24 = 1242.31; P = .030) (Table 4 and Figure).
The implementation of the COVID-19 survey within the MG’s PHP significantly increased the number of patients contacted prior to COVID-19 diagnosis from 5% (3/66) to 71% (43/61) (P < .05). Under the PHP, patients received a diagnosis of and were treated early for COVID-19 and chronic conditions. After initiation of COVID-19 surveys on April 6, hospitalizations due to COVID-19 declined significantly (49% vs 28%; P < .05) along with a significant decrease in mortality (9.1% vs 1.6%; P < .05) (eAppendix Table 9). A correlation was found between CTF referrals and overall mortality reduction at the MG compared with control patients in Florida over time (r = 0.93; P < .001) (eAppendix Figure 7).
The COVID-19 pandemic has proven to be one of the greatest health threats in generations. Any intervention to lower the rate of hospitalization, morbidity, and mortality would clearly be beneficial. Our retrospective study showed that in a large managed care population, it was possible to decrease COVID-19 mortality using a comprehensive PHP. Moreover, outpatient treatment, presumably in the early phase of the disease, was associated with decreased hospitalizations. Given that patients were age- and gender-matched and were treated in the same counties, the only difference in their care is their enrollment in a PHP, which included COVID-19 risk stratification, treatment of chronic medical conditions, COVID-19 treatment protocols, and intensive monitoring.
Although the risk factors for COVID-19 have been studied in the general population, our first research question sought to identify risk factors in the managed care population. Our data showed similar factors for COVID-19 incidence, hospitalization, and mortality in the managed care population as have been published in the general population.16,17 Specifically, older age, hypertension, and diabetes were associated with increased rates of COVID-19 diagnosis. Among patients who received a diagnosis of COVID-19, older age, obesity, hypertension, CKD, and CHF were associated with COVID-19 hospitalizations. For example, a previous study noted that 52% of hospitalized patients with COVID-19 were male (compared with 54% in our study) and that obesity was a significant risk factor (adjusted odds ratio, 1.9), which is comparable with the increased risk we found in our study for obese patients (odds ratio, 2.8).18
In our second research question, we explored the clinical impact of off-label medications in the outpatient setting. To our knowledge, no study has explored this question, particularly in a high-risk managed care population. We found that the use of outpatient off-label treatment reduced hospitalizations (67% vs 24%; P < .05) but not COVID-19 mortality. However, none of the off-label medications used have been definitively proven to confer benefit in randomized studies.19 Because blood pressure, glucose control, and other factors were not measured for both groups, those receiving outpatient off-label treatment may have been better optimized and thus the treatment of chronic conditions represents a possible confounder.
Our final research question investigated the clinical outcomes of MCPs in the MG compared with an age- and gender-matched mirror group in the state of Florida. Although COVID-19 incidence and hospitalization rates were similar in the MG and in the state of Florida, mortality was 60% lower for all patients with COVID-19 and 63% lower for those who were hospitalized. Furthermore, the introduction of a COVID-19–specific PHP was highly correlated to that mortality reduction (r = 0.93; P < .001). We believe these beneficial results were due primarily to increased patient engagement (in the form of clinical encounters, home visits, and other communication) and medical optimization (such as evidence-based treatment of chronic conditions).
Previous research has shown that patient-centered engagement and intensive primary care may improve outcomes.8-10,20 Dunphy et al in a systematic literature review cited face-to-face interactions, risk stratification of populations, and disease-specific programs (among other best practices) to reduce hospital admissions and emergency department visits.21 Gray et al found that repeated contact between patients and doctors (referred to as “continuity of care”) was associated with lower mortality.12 Similarly, Jerant et al found that greater access to care was associated with increased survival.20
A study by Mays and Smith found that public health spending is linked to patient survival, at a rate of 1.1% to 6.9% for every 10% increase in local public health spending (which targets local health care needs, often through disease-specific screening, prevention, and treatment).22 Managed care, which incentivizes a comprehensive care approach to ensure best outcomes, has been linked to lower mortality for several medical conditions.23 Moreover, treatment of chronic conditions, such as hypertension and diabetes, may result in a reduction in COVID-19 mortality.24,25
Our results suggest that a comprehensive PHP, within the context of managed care, in which highest-risk patients are identified and treated early (specifically in the outpatient setting) may significantly reduce mortality from COVID-19. Although we used several off-label treatments, this study was not powered to evaluate the effect of any particular treatment modality. Moreover, the program included a number of interventions, including the delivery of pulse oximeters, which may have provided additional clinical benefit through selection of therapy and setting of care.26
Our study’s findings also suggest that a patient-centered approach to primary care may lead to significant improvements in health care outcomes, especially in the underserved population, who often lack the resources, awareness, or support to establish continuity of care. Further, our study adds to a growing body of work concluding that a health care culture that is removed from episodic care, particularly in a time of crisis such as COVID-19, can result in significant benefits.27 This is a critical narrative in low socioeconomic circumstances, where morbidity and mortality tend to be higher, with access to care representing a major impediment to improving health outcomes. COVID-19 has presented a monumental challenge to medicine and the health care system. The high incidence and complexity of this disease requires a paradigm shift in health care delivery to a system of care that proactively seeks to individualize care and maintain health across a population.
Our study was an outpatient retrospective cohort study. Inpatient treatment was not studied or compared. Mortality benefit in the MG may be due to a variety of factors, including the PHP, early arrival to the emergency department, patient education of COVID-19 perils, previous control of chronic conditions, treatment of exacerbations of chronic conditions, and other factors. In addition, the PHP’s protocol led to frequent contact and monitoring of the patient, which could have resulted in increased adherence to medications for chronic conditions. It should also be noted that there may be have been patients with COVID-19 who did not seek medical care (or otherwise communicate with their providers or state authorities) and thus were not counted among Florida- or MG-reported data. Moreover, there may have also been a lag in reporting due to high caseloads and other factors faced by some providers and institutions.
The MG’s treatment protocols changed during and after the study period to reflect emerging data and professional guidance. Given the improved understanding of COVID-19 during the study period, there may have been a progressive decrease in mortality (particularly at the inpatient level); however, this change would have similarly affected both the MG population and the Florida mirror group population. A prospective randomized trial would be required to lend further support to the benefits of a comprehensive PHP for MCPs, as presented in this manuscript.
The COVID-19 pandemic has been a stress test of our health care system in the most extreme of ways. This study reviewed more than 30,000 managed care patients treated by a value-based primary care group. Our data show that a proactive and coordinated approach to primary care improved outcomes, even when traditional access to care was limited or discouraged, and the population studied was among the most vulnerable to illness. These results argue for a greater emphasis on multifaceted primary care and population health programs rather than on any single type of intervention or place of service. They also highlight the importance of risk stratification and targeted prevention, particularly in the most complex populations, and therefore may represent a paradigm shift for US health care.
The authors wish to thank Cano Health clinicians, whose dedication to patient care is unmatched, and Kathleen Hagen, EdD, of Nova Southeastern University, for her help with drafting the manuscript. They also thank Bianca Barrow, Pedro Cordero, Toni Mikell, and the Cano Health Population Health and Care Management teams, who recorded detailed patient-level data and coordinated the COVID-19 Task Force program. Without them, this manuscript would not have been possible.
Author Affiliations: Cano Health (RBA, JM, MO, MBH), Miami, FL; Dr Kiran C. Patel College of Allopathic Medicine, Nova Southeastern University (PH, BSM), Fort Lauderdale, FL; Westside Regional Medical Center (GT), Plantation, FL; Cleveland Clinic Florida (DS), Weston, FL.
Source of Funding: None.
Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (RBA, DS, JM, MBH); acquisition of data (RBA, JM, MO, MBH); analysis and interpretation of data (RBA, PH, BSM, GT, JM, MBH); drafting of the manuscript (RBA, PH, BSM, GT, DS, JM, MBH); critical revision of the manuscript for important intellectual content (RBA, PH, BSM, GT, DS, JM, MO, MBH); statistical analysis (PH); provision of patients or study materials (RBA, MO, MBH); obtaining funding (RBA, MBH); administrative, technical, or logistic support (RBA, JM, MO, MBH); and supervision (RBA, MBH).
Address Correspondence to: Marlow B. Hernandez, DO, MPH, Cano Health, 9725 NW 117 Ave, Suite 200, Miami, FL 33178. Email: firstname.lastname@example.org.
1. Banerjee A, Pasea L, Harris S, et al. Estimating excess 1-year mortality associated with the COVID-19 pandemic according to underlying conditions and age: a population-based cohort study. Lancet. 2020;395(10238):1715-1725. doi:10.1016/S0140-6736(20)30854-0
2. Stokes EK, Zambrano LD, Anderson KN, et al. Coronavirus disease 2019 case surveillance—United States, January 22-May 30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69(24):759-765. doi:10.15585/mmwr.mm6924e2
3. Physician’s Briefing Staff. 22 percent worldwide at increased risk for severe COVID-19. HealthDay. June 19, 2020. Accessed June 21, 2020. https://www.consumer.healthday.com/infectious-disease-8/coronavirus-1008/22-percent-worldwide-at-increased-risk-for-severe-covid-19-758643.html
4. Liu Y, Mao B, Liang S, et al; Shanghai Clinical Treatment Experts Group for COVID-19. Association between age and clinical characteristics and outcomes of COVID-19. Eur Respir J. 2020;55(5):2001112. doi:10.1183/13993003.01112-2020
5. Yamada T, Chen CC, Murata C, et al. Access disparity and health inequality of the elderly: unmet needs and delayed healthcare. Int J Environ Res Public Health. 2015;12(2):1745-1772. doi:10.3390/ijerph120201745
6. Aguilar RB, Hardigan P, Mayi B, et al. Current understanding of COVID-19 clinical course and investigational treatments. Front Med (Lausanne). 2020;7:555301. doi:10.3389/fmed.2020.555301
7. Shi L. The impact of primary care: a focused review. Scientifica (Cairo). 2012;2012:432892. doi:10.6064/2012/432892
8. Wensing M, Szecsenyi J, Kaufmann-Kolle P, Laux G. Strong primary care and patients’ survival. Sci Rep. 2019;9(1):10859. doi:10.1038/s41598-019-47344-9
9. Franks P, Fiscella K. Primary care physicians and specialists as personal physicians: health care expenditures and mortality experience. J Fam Pract. 1998;47(2):105-109.
10. Pereira Gray DJ, Sidaway-Lee K, White E, Thorne A, Evans PH. Continuity of care with doctors—a matter of life and death? a systematic review of continuity of care and mortality. BMJ Open. 2018;8(6):e021161. doi:10.1136/bmjopen-2017-021161
11. Reis HT, Clark MS, Pereira Gray DJ, et al. Measuring responsiveness in the therapeutic relationship: a patient perspective. Basic Appl Soc Psych. 2008;30(4):339-348. doi:10.1080/01973530802502275
12. Gray DP, Evans P, Sweeney K, et al. Towards a theory of continuity of care. J R Soc Med. 2003;96(4):160-166. doi:10.1258/jrsm.96.4.160
13. Florida COVID-19 Response. Florida Department of Health. Accessed June 3, 2020. https://floridahealthcovid19.gov/
14. Chronic conditions. CMS. Accessed June 15, 2020. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main
15. Chapel JM, Ritchey MD, Zhang D, Wang G. Prevalence and medical costs of chronic diseases among adult Medicaid beneficiaries. Am J Prev Med. 2017;53(6 suppl 2):S143-S154. doi:10.1016/j.amepre.2017.07.019
16. Zheng Z, Peng F, Xu B, et al. Risk factors of critical & mortal COVID-19 cases: a systematic literature review and meta-analysis. J Infect. 2020;81(2):e16-e25. doi:10.1016/j.jinf.2020.04.021
17. Richardson S, Hirsch JS, Narasimhan M, et al. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA. 2020;323(20):2052-2059. doi:10.1001/jama.2020.6775
18. Killerby ME, Link-Gelles R, Haight SC, et al; CDC COVID-19 Response Clinical Team. Characteristics associated with hospitalization among patients with COVID-19—metropolitan Atlanta, Georgia, March-April 2020. MMWR Morb Mortal Wkly Rep. 2020;69(25):790-794. doi:10.15585/mmwr.mm6925e1
19. Horby P, Mafham M, Linsell L, et al. Effect of hydroxychloroquine in hospitalized patients with COVID-19: preliminary results from a multi-centre, randomized, controlled trial. medRxiv. Preprint posted online July 15, 2020. doi:10.1101/2020.07.15.20151852
20. Jerant A, Fenton J, Franks P. Primary care attributes and mortality: a national person-level study. Ann Fam Med. 2012;10(1):34-41. doi:10.1370/afm.1314
21. Dunphy JN, Weinberger M, Silberman P. Strategies for implementing best practices in independent physician associations. Am J Manag Care. 2020;26(6):262-266. doi:10.37765/ajmc.2020.43489
22. Mays GP, Smith SA. Evidence links increases in public health spending to declines in preventable deaths. Health Aff (Millwood). 2011;30(8):1585-1593. doi:10.1377/hlthaff.2011.0196
23. Rogowski J, Jain AK, Escarce JJ. Hospital competition, managed care, and mortality after hospitalization for medical conditions in California. Health Serv Res. 2007;42(2):682-705. doi:10.1111/j.1475-6773.2006.00631.x
24. Gao C, Cai Y, Zhang K, et al. Association of hypertension and antihypertensive treatment with COVID-19 mortality: a retrospective observational study. Eur Heart J. 2020;41(22):2058-2066. doi:10.1093/eurheartj/ehaa433
25. Wang S, Ma P, Zhang S, et al. Fasting blood glucose at admission is an independent predictor for 28-day mortality in patients with COVID-19 without previous diagnosis of diabetes: a multi-centre retrospective study. Diabetologia. 2020;63(10):2102-2111. doi:10.1007/s00125-020-05209-1
26. Shah S, Majmudar K, Stein A, et al. Novel use of home pulse oximetry monitoring in COVID-19 patients discharged from the emergency department identifies need for hospitalization. Acad Emerg Med. 2020;27(8):681-692. doi:10.1111/acem.14053
27. Bakshi S, Schiavoni K, Carlson LC, et al. The essential role of population health during and beyond COVID-19. Am J Manag Care. 2021;27(3):123-128. doi:10.37765/ajmc.2021.88511