Predictors of Adverse Outcomes in Hospitalized Adult Patients with Sickle Cell Disease.

The identification of patients with sickle cell disease at risk of serious complications at the time of hospital admission can help stratify patients who will need aggressive management. We identified predictors associated with adverse outcomes such as frequent hospitalizations, acute pain crises and acute chest syndromes. To that end, we retrospectively reviewed medical records of 265 adult sickle cell disease patients, hospitalized between 1/1/98 and 2/3/05 at Mercy Catholic Medical Center, with complete clinical and laboratory data. 195/73.6% had HbSS and the rest had HbSC, HbSβ-^thal0,HbS-β^thal+or HbSO^arab disease. 59 variables were considered including demographic, hematological, biochemical, clinical and treatment data. Logistic regression models were used to obtain associations between variables, and to adjust for confounding effects. Analysis showed that adverse events during admission included acute pain crises in 249/94%, acute chest syndromes in 25/9.4% and strokes in 5/1.9% patients. Other outcomes were a greater than 2 hospitalizations per year 82/31.9%, more than 2 pain crises per year 145/54.7%, transfusion required during admission 72/27.2%, length of hospital stay more than 5 days 105/39.6% and death during hospitalization 13/4.9%. Multivariate logistic regression analysis revealed 21 factors with statistically significant associations. A reticulocyte count greater than 1.5 (OR 3.98, CI 1.48–10.69, P.006) and employment status (OR .31, CI .13-.75, P.009) were associated with more admissions per year. History of acute chest syndrome (OR 5.33, CI 1.7–16.77, P.004), reticulocyte count greater than 1.5 (OR 3.46, CI .91–13.11, P.067) and care provided by a nonhematologist (OR 5.04, CI 1.7–14.95, P.0035) were linked with more pain crises per year. Pain crises during admission were associated with HbSS disease (OR 9.31, CI 2.17–39.9, P.01) and out patient folate therapy(OR 6.23, CI 1.45–26.84, P.003). Patients with leukocytosis (OR 3.41, CI 1.2–9.67, P.02) and a higher serum glucose level (OR 7.54, CI 2.6–21.86, P.0002) were linked to more acute chest syndromes. Females (OR .1, CI .03–.37, P.0004) were at lower risk of having acute chest syndromes. Outpatient folate therapy (OR .07, CI .007–.69, P.02) was associated with lower numbers of acute neurological events. Patients with initial hemoglobin levels less than 7 g/dL (OR 1.99, CI 1–4, P.0007) and prolonged hospitalization (OR 7.06, CI 3.63–13.74, P.0001) frequently required transfusions. Variant diseases (OR .28, CI .13–.58, P.05) required fewer transfusions. Deaths during hospitalization were lower with folate therapy (OR .18, CI .05–.63, P.007) and a transfusion requirement during admission (OR 5.07, CI 1.45–17.64, P.01) predicted more deaths. HbSS patients (OR 2.52, CI 1.1–5.8, P.03), substance abusers (OR 2.93, CI 1.21–7.08, P.01), those requiring antihistamines during admission (OR 3.33, CI 1.38–8.03, P.007), or requiring more than 2 hospitalizations per year (OR 2.62, CI 1.26–5.43, P.009) had hospital stays longer than 5 days while in females odds were low for this outcome (OR .30, CI .15–.59, P.0005). In conclusion, simple tools like a complete history, physical examination, demographic and laboratory data can help clinicians and health care providers to gauge severity of the illness and deliver tailored management protocols targeting these “at risk” sickle cell disease patients.