Cost Drivers Associated with Clostridium difficile-Associated Diarrhea in a Hospital Setting

Journal of Clinical Outcomes Management. 2015 March;March 2015, VOL. 22, NO. 3:

March 13, 2015|Journal of Clinical Outcomes Management

Swetha Rao Palli, MS;Kelly C. Broderick, PharmD;Ralph A. Quimbo, MA;Marcie E. Strauss, MPH

Methods

Population

Patients were selected for this retrospective medical record review from the HealthCore Integrated Research Database (HealthCore, Wilmington, DE). The database contains a broad, clinically rich and geographically diverse spectrum of longitudinal claims information from one of the largest commercially insured populations in the United States, representing 48 million lives. We identified 21,177 adult (≥ 18 years) patients with at least 1 inpatient claim with an International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM) diagnosis code for C. difficile infection (CDI; 008.45) between 1 January 2005 and 31 October 2010 (intake period). All patients had at least 12 months of prior and continuous medical and pharmacy health plan eligibility prior to the incident CDAD-associated hospitalization within the database. Additional details regarding this cohort identification has been published previously [11]. The study was undertaken in accordance with Health Insurance Portability and Accountability Act (HIPAA) guidelines and the necessary central institutional review board approval was obtained prior to medical record identification and abstraction.

Sampling Strategy

To achieve the target sample of 500 fully abstracted medical records, 2500 patients were randomly selected via systematic random sampling without replacement from the original CDAD population of 21,177; their earliest hospital stay during the intake period with a diagnosis of CDI was identified using medical claims’ data and targeted for medical record abstraction. To be considered eligible for abstraction, medical records were required to include physicians’ and nurses’ notes/reports, discharge summary notes, medication administration record (MAR), and confirmation of CDAD via a documented CDI ICD-9-CM diagnosis or written note by a physician or nurse. Records with invalid or missing hospital or patient names were deemed ineligible for abstraction. Medical record retrieval continued until the requisite 500 CDAD-validated medical records were abstracted (Figure).

Medical Record Abstraction

During the record abstraction process, information was collected on patients’ race/ethnicity, body mass index (BMI), admission diagnosis and other conditions, point of entry and prior location, body temperature and laboratory data (eg, creatinine and albumin values, white blood cell [WBC] count), diarrhea and stool characteristics, CDAD diagnosis date, CDAD-specific characteristics, severity, complications, and related tests/procedures, CDAD treatments (eg, dose, duration, and formulation of medications), hospital LOS, including stays in the intensive care unit (ICU), cardiac care unit (CCU) following CDAD diagnosis; consultations provided by gastrointestinal, infectious disease, intensivists, or surgery care specialists, and discharge summary on disposition, CDAD status, and medications prescribed. Standardized data collection forms were used by trained nurses or pharmacists to collect information from the medical records and inter-rater reliability testing with a 0.9 cutoff was required to confirm accuracy. To ensure consistency, a pilot test of the first 20 abstracted records were re-abstracted by the research team. Last, quality checks were implemented throughout the abstraction process to identify any inconsistencies or data entry errors including coding errors and atypical, unrealistic data entry patterns (eg, identical values for a particular data field entered on multiple records; implausible or erratic inputs; or a high percentage of missing data points). Missing data were not imputed.

Study Definitions

Diarrhea was defined as 3 or more unformed (includes bloody, watery, thin, loose, soft, and/or unformed stool) bowel movements per day.CDAD severity was classified as mild (4–5 unformed bowel movements per day or WBC ≤ 2000/mm³); moderate (6–9 unformed bowel movements per day or WBC between 12,001/mm³ and 15,000/mm³); or severe (≥10 unformed bowel movements per day or WBC ≥15,001/mm³) [12,13]. Diarrhea was considered to be resolved when the patient had no more than 3 unformed stools for 2 consecutive days and lasting until treatment was completed, with no additional therapy required for CDAD as of the second day after the end of the course of therapy [2,14].CDAD episode was defined as the duration from the date of the CDAD diagnosis or confirmation (whichever occurred first), to the date of diarrhea resolution (where documented) or discharge date.

Cost Measures

The total hospital health plan paid costs for the entire inpatient episode (includes treatment costs, diagnostics, services provided, etc.) were estimated using medical claims present in the database and pertaining to the hospitalization from where medical records were abstracted. Then the proportionate amount for the duration of the CDAD episode (from CDAD diagnosis to the diarrhea resolution date or the discharge date in cases where the resolution date could not be ascertained) was calculated to estimate the average CDAD associated in-hospital costs.

Analysis

Means (± standard deviation [SD]), medians (interquartile range Q1 to Q3), and relative frequencies were calculated for continuous and categorical data, respectively. This analysis was descriptive in nature; hence, no statistical tests to determine significance were conducted.

References