Predictors of suboptimal and crash initiation of dialysis at two tertiary care centers

Many end-stage renal disease patients do not have an optimal start to dialysis. Many patients have suboptimal initiation, while others "crash" start on dialysis without prior care from a nephrologist. We examined factors associated with suboptimal or crash starts. We conducted a retrospective cohort study of 377 incident dialysis patients at two tertiary care centers from January 2006 to April 2011. Logistic regression was used to identify factors associated with suboptimal and crash starts to dialysis. Out of 377 patients, 102 (27%) had optimal starts, 221 (59%) had suboptimal starts, and 54 (14%) had crash starts. Three hundred thirty-four patients (89%) began with hemodialysis, while 11% started with peritoneal dialysis. Factors independently associated with a suboptimal start as opposed to an optimal start included nephrology care more than 12 months prior to initiation of dialysis (odds ratio [OR], 0.26; 95% confidence interval [CI], 0.12-0.58), Charlson Comorbidity Index (OR, 1.25 per 1 point; 95% CI, 1.09-1.43), and age (OR, 1.02 per 1 year; 95% CI, 1.00-1.04). In comparison, diabetic nephropathy (OR, 0.25; 95% CI, 0.12-0.54), a history of pulmonary edema within 6 months prior to initiation of dialysis (OR, 3.70; 95% CI, 1.77-7.75), and a diagnosis of chronic obstructive lung disease (OR, 0.07; 95% CI, 0.01-0.52) were independently associated with a crash start. There was a low incidence of optimal dialysis starts in our tertiary care dialysis population. Our study highlights that suboptimal and crash start patients are distinct populations. Modifying factors that predict nonoptimal dialysis starts will need to consider these distinctions.