The relationship of PROMIS physical function scores and healthcare resource utilization in patients treated for chronic low back pain

Abstract

Background context

Patients with mechanical chronic lower back pain (CLBP) have few durable treatment options for their condition and thus suffer decreased productivity and have higher healthcare resource utilization (HRU) compared to patients without CLBP. The economic burden of treatment and ongoing care for CLBP is considerable, with healthcare spending in 2016 estimated at $134.5 billion in the United States.

Purpose

This study aims to assess the correlation between patient-reported physical function scores and HRU in patients treated for mechanical CLBP.

Study design

This was a retrospective cohort study within a university-based health system.

Patient sample

Patients with a diagnosis of mechanical CLBP from 2015 through 2020 (index date) who were non-surgical candidates at baseline were included in this study. To ensure the presence of chronic low back pain, patients were required to have encounters between 6 and 12 months as well as between 12 and 24 months following the date of CLBP diagnosis.

Outcome measures

Collected data variables included patient baseline characteristics, Patient-Reported Outcomes Measurement Information System - Physical Function (PROMIS-PF) scores, pharmacologic and non-pharmacologic therapies, HRU, and healthcare charges between January 2015 through December 2022.

Methods

PROMIS-PF scores were converted to numerical categories ranging from 0 to 3, with Category 0 representing the lowest physical function and Category 3 the highest physical function. Patients were more broadly stratified into Low Physical Function (Low-PF) (Category 0–1) or High-Physical Function(High-PF) (Category 2–3) cohorts. HRU was compared between the Low-PF and High-PF cohorts using linear regression analyses. A mixed-effects regression analysis comparing Low-PF and High-PF patients was performed to model the relationship between patient-reported physical function and healthcare charges. The model is able to estimate charges for a base-case patient and can be adjusted to include patient-specific characteristics.

Results

A total of 2765 patients were included in this study, mean age was 50.1 (SD:17.7) years old, 23.6 % were 65 years or older, 68.4 % were female, and 85.3 % were white. Median healthcare charges by PROMIS-PF categories for Year-1 were highest for Category 0 patients ($14,028 [IQR: $5190-38,289]) and lowest for Category 3 ($5352 [IQR: $2417-14,470]). Patients in the Low-PF cohort showed significantly higher rates of all-cause, inpatient stays, outpatient visits, and emergency department (ED) visits compared to High-PF patients. The mixed effects regression model estimated cumulative healthcare charges to be > 2-fold higher for a base-case patient in the Low-PF cohort compared to High-PF. A small portion of patients (n = 14) failed treatment strategies and went on to receive CLBP-surgery despite not having surgical indications at baseline. Median healthcare charges from the 3-month period surrounding date of surgery were $59,809 (IQR: $46,057–85,484).

Conclusions

Cumulative Year-1 healthcare charges were almost 3-fold higher in Low-PF patients compared to High-PF. The mixed effects regression model estimated cumulative 2-year charges to be over 2-fold higher for Low-PF compared to High-PF, in the base-case patient. There were significantly higher rates of all-cause inpatient, outpatient, and ED visits in the year following diagnosis of mechanical CLBP for Low-PF patients. Despite receiving treatment, some patients went on to receive costly surgical procedures over the course of follow-up.