Evaluating How Safety-Net Hospitals Are Identified: Systematic Review and Recommendations

Utilization of Hip or Knee MRI in Patients 50 Years and Older With Atraumatic Pain: An Analysis of the National Ambulatory Medical Care Survey

PURPOSE

The aim of this study was to use the National Ambulatory Medical Care Survey database to assess MRI utilization in patients 50 years and older with atraumatic hip or knee pain.

METHODS

National Ambulatory Medical Care Survey weighted survey data (2007-2019) were obtained for ambulatory visits in patients 50 years and older with atraumatic hip or knee pain. The outcome variable was MRI ordering status, and analyzed characteristics included patient age, race/ethnicity, payer, physician specialty, metropolitan statistical area, and a coexistent radiography order. Multivariable logistic regressions were conducted to assess the association between MRI ordering status and the analyzed patient characteristics. All tests were two sided, and P values ≤.05 were considered to indicate statistical significance.

RESULTS

In total, 88,978,804 knee pain and 28,675,725 hip pain patient visits (survey weighted) were analyzed, with 4,690,943 (5.3%) and 2,023,226 (7.1%) having knee or hip MRI orders, respectively. Overall, 2,454,433 knee pain visits (2.8%) and 575,155 hip pain visits (2.0%) had orders for both MRI and radiographs. Black patients (P = .03) and patients 80 years and older (P = .04) were less likely to have knee MRI ordered, whereas uninsured patients were less likely to have hip MRI ordered (P = .01). Patients with hip pain were more likely to have hip MRI ordered if seen by a surgical subspecialist (P = .01).

CONCLUSIONS

A low proportion of MRI examinations were ordered for visits in patients 50 years and older with atraumatic hip or knee pain. Groups with lower health care access were less likely to have an MRI order, highlighting known disparities in health care equity.

Do Quality Measures or Hospital Characteristics Predict Readmission Penalties for Hip and Knee Arthroplasty?

BACKGROUND

Quality rating systems exist to grade the value of care provided by hospitals, but the extent to which these rating systems correlate with patient outcomes is unclear. The association of quality rating systems and hospital characteristics with excess readmission penalties for total hip arthroplasty (THA) and total knee arthroplasty (TKA) was studied.

METHODS

The fiscal year 2022 Inpatient Prospective Payment System final rule was used to identify 2,286 hospitals subject to the Hospital Readmissions Reduction Program. Overall, 6 hospital quality rating systems and 5 hospital characteristics were obtained. These factors were analyzed to determine the effect on hospital penalties for THA and TKA excess readmissions.

RESULTS

Hospitals that achieved a higher Medicare Overall Hospital Quality Star Rating demonstrated a significantly lower likelihood of receiving THA and TKA readmission penalties (Cramer's V = 0.236 and Rp = -0.233; P < .001 for both). Hospitals ranked among the US News & World Report's top 50 best hospitals for orthopaedics were significantly less likely to be penalized (V = 0.042; P = .043). The remaining 4 quality rating systems were not associated with readmission penalties. Penalization was more likely for hospitals with fewer THA and TKA discharges (Rp = -0.142; P < .001), medium-sized institutions (100 to 499 beds; V = 0.075; P = .002), teaching hospitals (V = 0.049; P = .019), and safety net hospitals (V = 0.043; P = .039). Penalization was less likely for West and Midwest hospitals (V = 0.112; P < .001).

CONCLUSIONS

A higher Overall Hospital Quality Star Rating and recognition among the US News & World Report's top 50 orthopaedic hospitals were associated with a reduced likelihood of THA and TKA readmission penalties. The other 4 widely accepted quality rating systems did not correlate with readmission penalties. Teaching and safety net hospitals may be biased toward higher readmission rates.

Safety-Net Hospital Status Is Associated With Coronary Artery Bypass Grafting Outcomes at an Urban Academic Medical Center

INTRODUCTION

Socioeconomic disparities impact outcomes after cardiac surgery. At our institution, cardiac surgery cases from the safety-net, county funded hospital (CH), which primarily provides care for underserved patients, are performed at the affiliated university hospital. We aimed to investigate the association of socioeconomic factors and CH referral status with outcomes after coronary artery bypass grafting (CABG).

METHODS

The institutional Adult Cardiac Surgery database was queried for perioperative and demographic data from patients who underwent isolated CABG between January 2014 and June 2020. The primary outcome was major adverse cardiovascular event (MACE), a composite of postoperative myocardial infarction, stroke, or death. Secondary outcomes included individual complications. Chi-square, Wilcoxon rank-sum, and logistic regression analyses were used to compare differences between CH and non-CH cohorts.

RESULTS

We included 836 patients with 472 (56.5%) from CH. Compared to the non-CH cohort, CH patients were younger, more likely to be Hispanic, non-English speaking, and be completely uninsured or require state-specific financial assistance. CH patients were more likely to have a history of tobacco and drug use, liver disease, diabetes, prior myocardial infarction, and greater degrees of left main coronary and left anterior descending artery stenosis. CH cases were less likely to be elective. The incidence of MACE was significantly higher in the CH cohort (16.3% versus 8.2%, P = 0.001). There were no significant differences in 30-d mortality, home discharge, prolonged mechanical ventilation, bleeding, sepsis, pneumonia, new dialysis requirement, cardiac arrest, or multiorgan system failure between cohorts. CH patients were more likely to develop renal failure and less likely to develop atrial fibrillation. On multivariable analysis, CH status (odds ratio 2.39, 95% confidence interval 1.25-4.55, P = 0.008) was independently associated with MACE.

CONCLUSIONS

CH patients undergoing CABG presented with greater comorbidity burden, more frequently required nonelective surgery, and are at significantly higher risk of postoperative MACE.

Care for Vulnerable Populations with Chronic Liver Disease: A Safety-Net Perspective

Safety-net hospitals (SNHs) and facilities are the cornerstone of healthcare services for the medically underserved. The burden of chronic liver disease-including end-stage manifestations of cirrhosis and liver cancer-is high and rising among populations living in poverty who primarily seek and receive care in safety-net settings. For many reasons related to social determinants of health, these individuals often present with delayed diagnoses and disease presentations, resulting in higher liver-related mortality. With recent state-based policy changes such as Medicaid expansion that impact access to insurance and critical health services, an overview of the body of literature on SNH care for chronic liver disease is timely and informative for the liver disease community. In this narrative review, we discuss controversies in the definition of a SNH and summarize the known disparities in the cascade of the care and management of common liver-related conditions: (1) steatotic liver disease, (2) liver cancer, (3) chronic viral hepatitis, and (4) cirrhosis and liver transplantation. In addition, we review the specific impact of Medicaid expansion on safety-net systems and liver disease outcomes and highlight effective provider- and system-level interventions. Lastly, we address remaining gaps and challenges to optimizing care for vulnerable populations with chronic liver disease in safety-net settings.

Comparison of Mirena and Liletta levonorgestrel intrauterine devices for the treatment of endometrial intraepithelial neoplasia and grade 1 endometrioid endometrial cancer

Objective

Current standard nonsurgical management of endometrial intraepithelial neoplasia (EIN) and grade 1 endometrioid endometrial cancer (g1EEC) is the Mirena levonorgestrel intrauterine device (M-IUD). This retrospective study was designed primarily to determine noninferiority of the Liletta IUD (L-IUD) for pathologic regression of EIN and g1EEC compared to the M-IUD at 6 months of continuous use. Secondary objectives include to determine noninferiority as above at 3, 9, and 12 months of continuous use and to identify factors including DNA mismatch repair (MMR) status associated with pathologic regression after LNG-IUD use.

Methods

A retrospective observational study was performed with patients treated for EIN or g1EEC and managed continuously with M- or L-IUD. Patients with recent (within 6 months) or concurrent progesterone use were excluded. For the EIN group, the noninferiority margin of odds ratio was predetermined to be 0.58, and for the g1EEC group it was 0.64.

Results

62 patients from an academic center and a safety-net hospital were identified with continuous M-IUD (n = 44) or L-IUD (n = 18) use for EIN or g1EEC. 85% of patients treated with L-IUD were from a safety-net hospital, which had 63% with public insurance. At 3/6/9 months, 54/71/73% of patients with M-IUD and 80/83/100% with L-IUD had pathologic regression of EIN (95% confidence interval of estimated odds ratio 1.00-2.07/0.84-2.03/0.69-2.10). Lifetime smoking status, not MMR status, was significantly associated with pathologic regression.

Conclusions

L-IUD is an effective fertility-sparing treatment for EIN. L-IUD is noninferior to M-IUD for pathologic regression of EIN after 3,6, and 9 months. Further larger studies are warranted to validate findings in EIN and g1EEC.

The intersection of drug use discrimination and racial discrimination in the management of chronic non-cancer pain in United States primary care safety-net clinics: Implications for healthcare system and clinic-level changes

BACKGROUND

Clinicians' bias related to patients' race and substance use history play a role in pain management. However, patients' or clinicians' understandings about discriminatory practices and the structural factors that contribute to and exacerbate these practices are underexamined. We report on perceptions of discrimination from the perspectives of patients with chronic non-cancer pain (CNCP) and a history of substance use and their clinicians within the structural landscape of reductions in opioid prescribing in the United States.

METHODS

We interviewed 46 clinicians and 94 patients, using semi-structured interview guides, from eight safety-net primary care clinics across the San Francisco Bay Area from 2013 to 2020. We used a modified grounded theory approach to code and analyze transcripts.

RESULTS

Clinicians discussed using opioid prescribing guidelines with the goals of increased opioid safety and reduced bias in patient monitoring. While patients acknowledged the validity of clinicians' concerns about opioid safety, they indicated that clinicians made assumptions about opioid misuse towards Black patients and patients suspected of substance use. Clinicians discussed evidence of discrimination in opioid prescribing at the clinic-wide level; racialized stereotypes about patients likely to misuse opioids; and their own struggles to overcome discriminatory practices regarding CNCP management.

CONCLUSION

While clinicians and patients acknowledged opioid safety concerns, the practical application of opioid prescribing guidelines impacted how patients perceived and engaged with CNCP care particularly for patients who are Black and/or report a history of substance use. We recommend healthcare system and clinic-level interventions that may remediate discriminatory practices and associated disparities.