Proving the Effectiveness of Supportive Care
Statistically Demonstrating the Effectiveness of Exercise and Nutrition Treatment for Cancer Patients
CLIENT | MAPLE TREE
Maple Tree Cancer Alliance, founded by Dr. Karen Wonders, is currently operating in eight hospitals serving Ohio and Pennsylvania. The organization is providing supportive care services to improve the quality of life for patients receiving cancer treatment. Through exercise programs and nutritional guidance, Maple Tree helps relieves some of the side effects related to various types of cancer treatment. This study sought to objectively validate the effectiveness of these programs and services.
ABOUT THE DATA
This study leverages data from GDAHA’s 360° Healthcare Database, which provides over 14 million comprehensive patient encounter observations across 25 regional hospital organizations for the time-period of January 2012 to September 2017. Using first name, last name, and birthdate, we matched 147 patients to corresponding GDAHA hospital records. Record search criteria aggregated encounters occurring six months prior and six months after the supportive care enrollment date for each patient. Therefore, one year of data was collected for each patient with enrollment occurring at the median. Due to our data ending in September 2017, the latest enrollment date for patients to be included in the study was March 31, 2017. The earliest enrollment date matched was November 11, 2014.
The resulting dataset consisted of 1,493 total hospital encounters for 147 unique patients; 60% of the patients had breast cancer. Most of the encounters were outpatient; of the 1,493, only 60 (4%) were inpatient. To demonstrate the effectiveness of supportive care, we focused on three primary measures for each patient: number of hospital encounters, readmissions and total charges. We also looked at length of stay and emergency room visits, which showed a decrease post-enrollment, but did not exhibit enough cases to be statistically relevant.
ABOUT THE PATIENTS
The primary objective of the analysis was to show a clear, statistical difference in hospital-related measures before and after enrollment in supportive care services provided by Maple Tree Cancer Alliance. To accomplish this, we split the data into pre and post-enrollment datasets (six months each), calculated the measures for both groups, and analyzed the differences. Visualizations and paired t-tests were used to demonstrate the impact of supportive care. A T-test is a statistical testing method often used to compare the before-and-after observations of repeated subjects. All assumptions for the t-tests were validated prior to generating results to ensure quality results (continuous dependent variable, independent observations, approximate normal distribution, and absence of outliers). Some measures such as length of stay and number of emergency room visits did not have a significant incidence rate, so only the change in values are reported.
LENGTH OF STAY & EMERGENCY ROOM VISITS
Length of Stay is typically calculated based on inpatient stays, which accounted for an extremely small percentage of the encounters included in the study. Despite the low numbers, we still found a 6% decrease in the number of inpatient stays and a 19% decrease in the length of stays associated with those admissions following supportive care enrollment. Length of stay nearly decreased by a full day within the post-enrollment encounters. A similar trend was found with the number of emergency room visits, which decreased by 27%.
STATISTICAL ANALYSIS OVERVIEW
The three primary hospital measures leveraged for statistical comparison before and after supportive care enrollment were number of encounters, number of readmissions, and average total charges. Here, we provide an overview of the statistical analysis for each measure; visualizations of the change in measures for each patient are provided in the next sections.
Number of encounters reflects each time a patient had an inpatient or outpatient encounter with a GDAHA hospital. In the data prior to patient enrollment, there were 898 total encounters. Following patient enrollment, the data showed 595 total encounters; a 33% decrease. The table shows that on average, each patient exhibited 2.06 less encounters following enrollment in supportive care; the t-test proved this to be a statistically significant decrease.
A readmission for this study is defined as a patient returning to a hospital within 30 days of a previous visit, regardless of inpatient or outpatient status. This does not include rules incorporated in the CMS definition of a readmission. We are purely measuring the frequency that a patient returns to the hospital. Prior to enrollment, there were 634 readmissions. Following enrollment, there were 351 readmissions; a 47% decrease. The table shows that on average, each patient had approximately two less readmissions across the board; the t-test proved this to be a statistically significant decrease.
Total charges were averaged across all encounters before and after enrollment to understand the average total charges per encounter; in this case this is the bill received by the payer before adjustments. The average cost per encounter decreased by $2,834 (28%) per patient after enrollment; the t-tested proved this to be a statistically significant decrease.
We found Maple Tree’s services to be effective in improving key hospital measures. Number of hospital encounters, readmissions, and average total charges exhibited statistically significant decreases (at least 30%) following patient enrollment in supportive care.