Impact of Weekend Admissions on Quality of Care and Outcomes in Patients with Acute Myeloid Leukemia (AML).

Background: Hospital services can be significantly reduced over the weekend. This was associated with higher mortality among patients with serious medical conditions in previous studies. It is unknown whether weekend admissions affect outcomes of AML patients, as delay in treatment or delay in obtainment of procedures (such as triple-lumen catheter [TLC] placement requisite for anthracycline delivery) may occur for weekend admissions. We investigated quality of care and clinical outcomes of newly-diagnosed AML patients treated with induction chemotherapy and hospitalized on weekends vs. weekdays.

Methods: We conducted a retrospective review of all AML patients treated with cytarabinebased induction chemotherapy at Cleveland Clinic from 1994–2008. Data on known prognostic factors (age, WBC at diagnosis, cytogenetic risk groups [as defined by CALGB 8461] and AML etiology [de novo vs. secondary AML]) were collected and controlled for in multivariable analyses. Quality measures included time to TLC placement; time to induction chemotherapy (TTI); length of stay (LOS); early death (within 15 days of chemotherapy initiation); and 30-day mortality. Weekend admissions were defined as starting Friday, 5pm through Monday, 12am. Factors associated with quality of care and outcomes were assessed by the routines of linear, categorical, and survival analyses.

Results: In all, 422 patients were identified. Median age was 61 years (range:17–81) and 47% were female. Median baseline WBC was 9.9/mcL (range:0.4–550); 4.7% had acute promyelocytic leukemia, and 30% had secondary AML. Cytogenetics risk distribution was: favorable (11.6%); intermediate (41.2%); adverse (24.2%); unknown/no growth (23%). In all, 24.4% (n=103) were admitted on the weekend. The complete remission (CR) rate was 66.6%; Median times to TLC was 2 days (range:0–27); TTI was 2 days (range:0–22); and LOS was 32 days (4–91). Early death rate was 3.1%, and 30-day mortality 10.4%. Compared to younger (<60 years) patients, older patients had higher 30-day mortality (14.7% vs. 5.8%, p=0.003), early death (4.9% vs. 1.1%, p=0.025), TTI rates (2.7 days vs. 2.1 days, p=0.02), but lower CR (62.9% vs. 77.8%, p=0.001) and overall survival (OS) rates (195 vs. 474 days, p<0.0001). In univariate analyses, time to TLC was delayed for weekend vs. weekday admissions (3.8 vs. 2.7 days, p<0.01), as was a trend for LOS (35.9 vs. 33.5 days, p=0.09). Surprisingly, weekend admissions had lower early mortality (0% vs. 4%, p=0.04) and 30-day mortality (3.9 % vs. 12.3%, p=0.02). There was no difference in TTI (2.5 vs. 2.4 days, p=0.6), CR rates (72% vs. 69%, p=0.6), or OS (258 vs. 273 days, p=0.2). In multivariable analysis, only time to TLC remained significantly longer for weekend admissions (p<0.001).

Conclusions: Weekend admissions significantly delayed placement of TLC, and possibly LOS, without affecting other quality parameters or patient survival. This is likely due to immediate initiation of peripheral chemotherapy with cytarabine even prior to the placement of TLC for infusion of anthracyclines, and emphasizes the need to start chemotherapy promptly in AML patients.