Visualizing complex healthcare disparities: proof of concept for representing a cyclical continuum of care model for a retrospective cohort of patients with musculoskeletal infections.

BACKGROUND: Care continuum models (also known as care cascade models) are used by researchers and health system planners to identify potential gaps or disparities in healthcare, but these models have limited applications to complex or chronic clinical conditions. Cyclical continuum models that integrate more complex clinical information and that are displayed using circular data visualization tools may help to overcome these limitations. We performed proof-of-concept cyclical continuum modeling for one such group of conditions-musculoskeletal infections-and assessed for racial and ethnic disparities across the complex care process related to these infections. METHODS: Cyclical continuum modeling was performed in a diverse, retrospective cohort of 1648 patients with musculoskeletal infections, including osteomyelitis, septic arthritis, and/or infectious myositis, in the University of New Mexico Health System. Logistic regression was used to estimate the relative odds of each element or outcome of care in the continuum. Results were visualized using circularized, map-like images depicting the continuum of care. RESULTS: Racial and ethnic disparities differed at various phases in the care process. Hispanic/Latinx patients had evidence of healthcare disparities across the continuum, including diabetes mellitus [odds ratio (OR) 2.04, 95% confidence interval (CI): 1.61, 2.60 compared to a white non-Hispanic reference category]; osteomyelitis (OR 1.28, 95% CI: 1.01, 1.63); and amputation (OR 1.48; 95% CI: 1.10, 2.00). Native American patients had evidence of disparities early in the continuum (diabetes mellitus OR 3.59, 95% CI: 2.63, 4.89; peripheral vascular disease OR 2.50; 95% CI: 1.45, 4.30; osteomyelitis OR 1.43; 95% CI: 1.05, 1.95) yet lower odds of later-stage complications (amputation OR 1.02; 95% CI: 0.69, 1.52). African American/Black non-Hispanic patients had higher odds of primary risk factors (diabetes mellitus OR 2.70; 95% CI: 1.41, 5.19; peripheral vascular disease OR 4.96; 95% CI: 2.06, 11.94) and later-stage outcomes (amputation OR 2.74; 95% CI: 1.38, 5.45) but not intervening, secondary risk factors (osteomyelitis OR 0.79; 95% CI: 0.42, 1.48). CONCLUSIONS: By identifying different structural and clinical barriers to care that may be experienced by groups of patients interacting with the healthcare system, cyclical continuum modeling may be useful for the study of healthcare disparities.

Racial and ethnic disparities in surgical amputations following serious musculoskeletal infections in a diverse New Mexico cohort.

BACKGROUND: Patients with serious musculoskeletal infections may encounter health disparities across multiple phases of prevention and treatment, including surgical intervention. The purpose of this study was to identify and compare the predictors of surgical intervention and surgical amputation among patients with septic arthritis, osteomyelitis, and infectious myositis in a diverse cohort of patients from New Mexico. METHODS: A retrospective cohort from the University of New Mexico Health System was formed. Patients with septic arthritis, osteomyelitis, and/or infectious myositis who underwent surgical procedures or amputations were compared with those who did not, using predictive multivariable logistic regression modeling. The impact of diabetes mellitus (DM) as a predictor of surgical outcomes was evaluated. RESULTS: DM was a predictor of both surgical procedures and amputations in a diverse cohort of patients (n = 1694). Diabetes was more common in American Indian/Alaskan Native (AI/AN) patients. However, Black non-Hispanic/African American and Hispanic patients were more likely to undergo amputations, compared to AI/AN patients, even after adjustment for diabetes severity. CONCLUSIONS: Racial and ethnic disparities in infection-related amputation may differ from those observed for diabetes or for general access to surgical management. Interventions intended to prevent or treat serious musculoskeletal infections should consider health disparities that differ across the clinical care process.

A Cyclical Approach to Continuum Modeling: A Conceptual Model of Diabetic Foot Care.

"Cascade" or "continuum" models have been developed for a number of diseases and conditions. These models define the desired, successive steps in care for that disease or condition and depict the proportion of the population that has completed each step. These models may be used to compare care across subgroups or populations and to identify and evaluate interventions intended to improve outcomes on the population level. Previous cascade or continuum models have been limited by several factors. These models are best suited to processes with stepwise outcomes-such as screening, diagnosis, and treatment-with a single defined outcome (e.g., treatment or cure) for each member of the population. However, continuum modeling is not well developed for complex processes with non-sequential or recurring steps or those without singular outcomes. As shown here using the example of diabetic foot care, the concept of continuum modeling may be re-envisioned with a cyclical approach. Cyclical continuum modeling may permit incorporation of non-sequential and recurring steps into a single continuum, while recognizing the presence of multiple desirable outcomes within the population. Cyclical models may simultaneously represent the distribution of clinical severity and clinical resource use across a population, thereby extending the benefits of traditional continuum models to complex processes for which population-based monitoring is desired. The models may also support communication with other stakeholders in the process of care, including health care providers and patients.