Racial disparities in physician decision making for primary brain tumours

Prompt evaluation and early surgical resection of neurological tumours have been shown to improve long-term outcomes for patients with primary brain tumors.^1-4 Despite the high morbidity of neurological disease within the USA, there is little data considering physician decisions in the management of neurosurgical patients and the disparities that could arise from these decisions. Disparities work in this field has largely focused on patient factors such as access, income, insurance, geography, structural inequities, and operative and postoperative morbidity.^5,6 Resection is a standard treatment modality for those with primary brain tumours. The possibility that patients from minority ethnic backgrounds might have a decreased rate of recommendation and receipt of surgery has not been explored and could have a devastating effect on outcomes.

In this issue of The Lancet, John T Butterfield and colleagues⁷ sought to explore this question by using two national databases to investigate the status of racial or socioeconomic disparities in neurosurgical interventions for a primary brain tumour. The primary outcome focused on a surgeon’s recommendation for or against tumour resection for common primary brain tumours.

The Surveillance, Epidemiology, and End Results (SEER) Program database and the American College of Surgeons National Cancer Database (NCDB) were independently analysed to determine if an association between race and socioeconomic status and a recommendation against surgical resection exists. This was studied in the most common primary brain tumours—meningioma, glioblastoma, pituitary adenoma, astrocytoma, vestibular schwannoma, and oligodendroglioma. In the SEER database, patients with meningiomas (median age 70 years) had the highest percentage (59·5%) of receiving a recommendation against surgical resection, whereas those with oligodendrogliomas (median age 48 years) had the lowest percentage (13·2%). In the NCDB database, patients with vestibular schwannoma (median age 61 years) had the highest percentage (60·6%) of receiving a recommendation against surgical resection, whereas those with oligodendroglioma (median age 49 years) had the lowest percentage (12·9%).

A multivariable logistic regression analysis of the SEER database showed that Black race was independently associated with higher odds of recommendation against surgical resection compared with White race: meningioma (OR 1·20, 95% CI 1·12–1·27, p<0·0001), glioblastoma (1·16, 1·02–1·30, p=0·018), pituitary adenoma (1·19, 1·11–1·28, p<0·0001), and vestibular schwannoma (1·56, 1·26–1·93, p<0·0001). Similar findings were also found on multivariable regression analyses of the NCDB database: meningioma (OR 1·26, 95% CI 1·22–1·30, p<0·0001), glioblastoma (1·20, 1·13–1·29, p<0·0001), pituitary adenoma (1·26, 1·22–1·31, p<0·0001), vestibular schwannoma (1·30, 1·15–1·48, p<0·0001), and astrocytoma (1·29, 1·09–1·51, p=0·0021) populations.

This work by Butterfield and colleagues confirms what is well known—surgical inequities are created at multiple levels (patient, systemic, and provider factors).⁸ The strength of their study is found in the re-demonstration of results across two national registries. Importantly, the authors suggest that physician’s clinical decision making might not be made on purely objective medical data. This study’s notable limitation was that the SEER database does not include comorbidity information, although it was accounted for in the NCDB database, disallowing for this factor to be tested as a potential confounding variable.

Disparities arising from clinical decision making is a multifaceted issue and previous work has shown that there are six key themes that factor into these decisions: importance of race, patient-level issues, system-level issues, bias and racism, patient values, and communication.⁹ Butterfield and colleagues highlight the potential for implicit biases in decision making surrounding neurosurgical procedures. These biases contribute to health-care disparities by influencing and shaping physician behaviour and producing variations in medical management along the lines of race, ethnicity, gender, and other characteristics. Notably, implicit bias could arise when clinicians are not deliberate about considering other perspectives when providing patient care.¹⁰

The study by Butterfield and colleagues is novel in taking the first steps to highlight the disparity in physician decision making for primary brain tumours across the USA. Surgical inequities are a multifactorial issue, and this work could provide a signal for surgeons caring for these diseases to be cognisant of their biases in surgical assessments and for researchers to move beyond solely examining social factors that drive surgical outcomes. Future directions for this area of inquiry should consider the link of patient outcomes to surgical decision making, impact of physician reasoning and patient preference, and identify opportunities and interventions to address the multilevel barriers to receipt of equitable care in this disease base. It is imperative to move past description-based studies and seek to delve deeper on the clinician decision-making process and highlight solutions that lend to decreasing disparities in the pre-operative setting.