Describing Spine Surgery Patients by Healthcare Utilization Following Surgery

Introduction: Outcomes following spine surgery can be heterogeneous and difficult to track. Patient reported outcome measures are often used to described the success of surgery, but these measures have inherent drawbacks such as their subjectivity and the non-response bias from patients who fail to complete any surveys. Considering healthcare utilization in addition to these metrics offers a more objective, complete measure of how often a patient interacts with the healthcare system following surgery.

Methods: A retrospective cross-sectional study of all spine surgery patients at a single institution was performed. Any patient who underwent surgery between 2014 and 2020 and completed at least 3 months of follow-up with a PROMIS Physical Health survey score was included in the study. Healthcare utilization was calculated from a latent profile analysis of all healthcare utilization components (including hospital/office/phone encounters, EMR messages, imaging, pain medication prescriptions, and PT/OT referrals). The odds ratio for covariates of interest (age, BMI, pre-operative PROMIS score, and pre- to post-operative PROMIS score) was calculated with the high utilizers as the reference group.

Results: The study included 5602 patients sorted into four utilization categories: low utilizers (21.5%), moderate utilizers without advanced imaging (34.7%), moderate utilizers with advanced imaging (34.7%), and high utilizers (33.1%). Compared to high utilizers, low and moderate utilizers (without and with advanced imaging) were associated with better improvement in PROMIS-PH scores following surgery with 1.134 (95% CI 1.116-1.153, p < 0.05), 1.126 (95% CI 1.108-1.144, p < 0.05), and 1.064 (95% CI 1.048-1.079, p < 0.05) greater odds of having higher scores post-operatively, respectively. The probability of patients being categorized as low or moderate without advanced imaging utilizers increased as PROMIS-PH scores improved postoperatively.

Conclusion : By using an objective measure like healthcare utilizers, we demonstrated that patterns of utilization are able to identify both the sickest patients prior to surgery and the ones who have the poorest outcomes post-operatively. Following patients by their healthcare utilization would allow for easy patient monitoring after surgery and understanding these patterns may help to better stratify a patient’s likelihood of benefitting from surgical intervention.