Construct validation and comparison of a novel postoperative quality-of-life metric and the Short Form-36 in colorectal surgery patients

Purpose Several methods have been used to measure quality of life (QOL) in colorectal patients, but existing QOL assessment tools are often complex, require complicated analysis, lack specificity for colorectal surgery, and are not focused on assessment of perioperative care. The postoperative QOL (PQL) assessment is designed to capture subtle yet important QOL factors in an easy tool validated for the postoperative period. Although internally validated, PQL lacks external validation with a universally accepted QOL metric, such as the Rand Short Form-36 (SF-36). The purpose of this study was to externally validate the PQL metric to the SF-36 for colorectal surgery. Methods The PQL was designed using 14 questions ranked on a Likert scale (1-10) with surgeon and patient input. After obtaining institutional review board approval, 100 consecutive colorectal surgery patients at University Hospitals, Case Medical Center were administered baseline and postoperative (2, 4, 8, 12, 30, 60, and 90 days) PQL and SF-36 questionnaires prospectively. Patients undergoing colorectal surgery via an abdominal approach (laparoscopic or open) for benign or malignant disease were included. Factor analysis and Spearman's rank test were performed between each of the 8 SF-36 scales and the 14 PQL questions and summary score. Convergent validity was demonstrated using Spearman's correlation coefficient at the domain and scale levels. The degree of agreement between PQL and SF-36 was assessed through Bland-Altman plots. Pairwise comparisons were made to determine any significant differences between the 2 scales. Results Eighty-eight patients met all inclusion criteria and had a complete dataset, and were included in the analysis. SF-36 factor analysis confirmed comparability between the study group and the general population. All PQL items correlated with all 8 mental and physical health domains in the SF-36 (P <.0001). Bland-Altman plots demonstrated consistently similar measure for level of agreement between PQL and SF-36 as indicated by the 95% limits of agreement. Conclusion The PQL and SF-36 demonstrated a strong and consistent level of agreement across all 8 domains for pre- and postoperative scores in colorectal surgery patients. PQL is constructually valid in the perioperative period. Based on our analysis, the novel PQL metric represents a simple, point-of-care alternative to SF-36 for rapid QOL assessment, and validates use of the PQL metric in abdominal surgery.