Clinical and morphometric parameters of frailty for prediction of mortality following hepatopancreaticobiliary surgery in the elderly

Abstract

Background: Although frailty is a known determinant of poor postoperative outcomes, it can be difficult to identify in patients before surgery. The authors sought to develop a preoperative frailty risk model to predict mortality among patients aged 65 years or more.

Methods: Clinical and morphometric data including total psoas area (TPA), total psoas volume (TPV) and psoas density (Hounsfield unit average calculation, HUAC) were collected for patients undergoing hepatopancreaticobiliary (HPB) surgery between 2012 and 2014. Multivariable Cox proportional hazards regression was used to identify preoperative risk factors associated with 1-year mortality.

Results: The median age of the 518 patients included in the study was 72 (i.q.r. 68-76) years; 55·6 per cent of patients were men, and half of the cohort had multiple co-morbidities (Charlson co-morbidity index (CCI) of 4 or more, 55·6 per cent). TPA cut-offs to define sarcopenia were 552·7 mm(2) /m(2) in women and 702·9 mm(2) /m(2) in men; cut-offs for TPV were 18·2 cm(3) /m(2) in women and 26·2 cm(3) /m(2) in men, whereas HUAC cut-offs were 31·1 HU in women and 33·3 HU in men. The overall 1-year mortality rate was 14·1 per cent. In multivariable analysis, risk factors associated with 1-year mortality included CCI of 4 or above (hazard ratio (HR) 2·91, 95 per cent c.i. 1·47 to 5·77; P = 0·002), malignant disease (HR 3·94, 1·17 to 13·30; P = 0·027) and sarcopenia by HUAC (HR 1·85, 1·10 to 3·10; P = 0·021). A weighted 25-point composite score was developed to stratify patients at risk of 1-year postoperative mortality. The 1-year mortality rate was noted to be 2·5 per cent among patients scoring 0-10 (low risk), 17·3 per cent among patients scoring 11-20 (intermediate risk) and 29·2 per cent among those scoring between 21 and 25 (high risk) (P < 0·001).

Conclusion: Clinical and morphometric measures of frailty accurately predict the risk of 1-year mortality following HPB surgery in elderly patients, and can be used to risk-stratify patients appropriately.