Age, Comorbidity, Time on Chemotherapy Predict Increases in Hospitalizations, ER Visits Among CLL Patients

Researchers identified several risk factors leading to greater health-care resource use, including age, comorbidity burden, presence of adverse events, and duration of therapy, in patients with chronic lymphocytic leukemia (CLL), according to a poster presented at the 2015 ASH Meeting on Hematologic Malignancies.

Bruce Feinberg, DO, from Cardinal Health in Dublin, Ohio, and colleagues conducted a retrospective analysis of claims data looking for risk factors of emergency room (ER) visits and hospitalizations among patients with CLL to help guide management of these patients.

“Acute care interventions negatively impact patients’ quality of life and represent one of the major cost drivers in oncology care,” Dr. Feinberg and colleagues wrote. “Limited information exists in the literature on health-care resource utilization in patients with CLL throughout the course of their disease.”

To evaluate predictors for ER visits and hospitalizations in patients with CLL treated with anti-cancer systemic therapy, the investigators analyzed claims data from the MORE Registry. Patients treated with anti-cancer systemic therapy for CLL were retrospectively identified by ICD-9 codes (204.1, 204.10, 204.11, 204.12) during a 48-month period (August 2009 to August 2013), and selected for analysis of their available treatment history. Patients with secondary malignancies, pregnancy codes, and age <18 were excluded from this analysis.

A total of 2,013 patients were included in the study (39% female). Median age was 72 years, 41 percent of patients were aged ≥75 years. Patients tended to have a high comorbidity burden, according to scores on the Charlson Comorbidity Index. Mean score was 5.0; patients aged <65 years had a mean score of 3.4 and patients aged ≥65 years had a mean score of 5.6.

Thirty-four percent of the patients were treated in the relapse setting, 42 percent of patients had ER visits, and 39 percent of patients had been hospitalized.

Dr. Feinberg and co-authors performed univariate regression analysis to determine variables associated with an increase in ER visits and hospitalizations. A total of 22 variables were selected for analysis, based on expert medical opinion, for their potential to affect or predict ER visits and hospitalizations.

Of the 22 variables analyzed, 18 were significant predictors of hospitalizations and 16 were significant predictors of ER visits based on univariate analysis (TABLE).

As seen in the TABLE, the factors conferring the highest risk of health-care resource use were: presence of disease- or treatment-related adverse events (including anemia, thrombocytopenia, and infection), relapsed setting treatment, a greater number of rituximab maintenance lines, and use of supportive care.

“The majority of variables associated with an increase in ER visits were similar to the variables associated with an increase in hospitalizations,” the authors noted. Notably, they did not find any significant variables associated with a decrease in hospitalizations or ER visits.

“These data warrant consideration of age- and comorbidity-adjusted treatment choice in CLL patients who need active treatment with anti-cancer systemic therapies and effective management of adverse events,” they concluded.

However, there are important limitations to note.

First, this was a retrospective analysis of a payer claims database, and therefore “subject to inaccuracies in billing and missing data.” In addition, claims data did not include information such as disease severity or reasons for health-care resource use. Causal association, then, cannot be made between the variables assessed and health-care resource use. “However, comorbidities and time on anti-cancer systemic treatment can be used to illustrate more complex or advanced disease,” they wrote.

“The set of variables selected for analysis to assess whether they predict increased or decreased health-care resource utilization were based on expert medical opinion,” Dr. Feinberg and co-authors wrote. “Additional variables, especially those not available in claims data (e.g., clinical outcomes), may be explored in future studies.”

Reference

Feinberg B, Schenkel B, McBride A, et al. Predictors of increase in emergency room (ER) visits and hospitalizations in patients with chronic lymphocytic leukemia (CLL) treated with chemotherapy.  Abstract #30. Presented at the 2015 ASH Meeting on Hematologic Malignancies; September 19, 2015; Chicago, IL.

TABLE. Univariate Regression Analysis of Predictors of Hospitalizations and ER Visits in CLL Patients Treated With Anti-Cancer Systemic Therapies
Hospitalization ER Visit
Univariate odds ratio p Value Univariate odds ratio p Value
Age (continuous variable) 1.001 0.038 1.015 0.001
Treatment in the U.S. Northeast region 1.321 0.005 1.293 0.008
Charlson comorbidity index >0 1.268 0.010 1.291 0.005
Number of lines of anti-cancer systemic therapy 2.095 <0.0001 1.880 <0.0001
Number of rituximab maintenance lines 3.322 0.022 1.791 0.205
Any use of rituximab maintenance 1.899 <0.0001 1.470 0.003
# of CLL disease- or treatment-related AEs (continuous variable) 1.615 <0.0001 1.520 <0.0001
Presence of CLL disease- or treatment-related AEs 29.467 <0.0001 12.311 <0.0001
Nausea and vomiting presence 1.837 <0.0001 1.992 <0.0001
Neutropenia presence 1.944 <0.0001 1.668 <0.0001
Infection presence 4.610 <0.0001 4.487 <0.0001
Anemia presence 4.634 <0.0001 3.782 <0.0001
Thrombocytopenia presence 4.253 <0.0001 3.630 <0.0001
Number of supportive care agents used 1.278 <0.0001 1.262 <0.0001
Any supportive care use 2.429 <0.0001 2.087 <0.0001
Total time on therapy from start of 1st line to last claim date (6-month increments) 1.268 <0.0001 1.292 <0.0001
Treatment following relapse 4.018 <0.0001 3.413 <0.0001
Duration of anti-cancer systemic therapy (6-month increments) 1.247 <0.0001 1.252 <0.0001

 

SHARE