Long-Term Consequences of Worsened Poststroke Status in Patients With Premorbid Disability

Abstract

Background and Purpose— Patients with premorbid disability, generally defined as modified Rankin Scale (mRS) score ≥2, are often excluded from trials of acute stroke therapies. However, increased disability in such patients will adversely affect long-term outcomes if treatments are withheld in routine practice. We assessed the extent to which increased disability poststroke influences 5-year mortality, institutionalization, and costs in premorbidly disabled patients. Methods— In a population-based, prospective cohort of patients with ischemic stroke (OXVASC [Oxford Vascular Study], 2002–2014), we tracked mortality, institutionalization, and healthcare/social-care costs during follow-up. We compared 5-year mortality and poststroke institutionalization (Cox regressions) and 5-year healthcare/social-care costs (generalized linear model) in 3-month survivors with premorbid mRS of 2 to 4 (excluding extreme disability, mRS=5), based on the degree of change in mRS(ΔmRS) from prestroke to 3 months poststroke, adjusting analyses for age/sex/initial National Institutes of Health Stroke Scale. Results— Among 1607 patients, 530 (33.0%) had premorbid mRS of 2 to 4. Only 2 premorbidly disabled patients received thrombolysis, but 421 (79.4%) were alive at 3 months. ΔmRS was independently associated with 5-year mortality/institutionalization (adjusted hazard ratio for ΔmRS=1 versus 0: 1.59; 95% CI, 1.20–2.11; ΔmRS=2: 2.39; 95% CI, 1.62–3.53; ΔmRS=3: 4.12; 95% CI, 1.98–8.60; P <0.001) and costs (margin for ΔmRS ≥2 versus 0: $30 011, 95% CI, $4222–55 801; P =0.023). Results were similar on examining patients with premorbid mRS of 2, 3, and 4 separately (eg, 5-year mortality/institutionalization adjusted hazard ratio for premorbid mRS=3 with ΔmRS=1 versus 0: 1.60; 95% CI, 1.06–2.42; P =0.027; ΔmRS=2: 3.20; 95% CI, 1.85–5.54; P <0.001). Conclusions— Patients with stroke with premorbid disability have higher mortality, institutionalization, and costs if they accumulate additional disability because of the stroke. These findings highlight the long-term outcomes expected if acute interventions are routinely withheld in patients with mild-moderate premorbid disability and suggest that trials/registries should include such patients.