Prioritized Post-Discharge Telephonic Outreach Reduces Hospital Readmissions for Select High-Risk Patients
Published Online: December 18, 2012
L. Doug Melton, PhD, MPH; Charles Foreman, MD; Eileen Scott, RN; Matthew McGinnis, BS; and Michael Cousins, PhD
Readmissions make up a significant portion of healthcare expenditures and account for 30% of total inpatient cost in the commercial population.1 The Agency for Healthcare Research and Quality stated that roughly 1 in 10 readmissions in 2008 could have been prevented if acute conditions were managed properly post discharge.2
Prior published studies testing the impact of post-discharge outreach on readmission reductions have shown success with select interventions such as Telehealth video consultations with an in-person health practitioner3 and house calls.4,5 In comparison, the evidence is mixed for the effectiveness of case management (CM) that is not in person.6-8 Riegel and colleagues tested the effect of telephonic CM for heart failure patients’ resource use and found the intervention group had 36% lower heart failure readmissions at 3 months than controls (P <.05). In contrast, Laramee tested the effect of telephonic CM paired with early discharge planning for congestive heart failure patients on readmission, but found 90-day readmission rates were equal for the CM and usual care groups. Neither of these studies focused solely on the impact of CM on the commercially insured. To date, there is only 1 published prospective, stratified randomized study assessing the impact of telephonic CM on readmissions within a large, commercially insured population, and it showed a 10% readmission reduction.9
The study objective was to determine if a telephonic CM patientprioritization protocol for recently discharged high-risk patients with select acute conditions reduced all-cause readmissions. We conducted a prospective, stratified randomized study to test the effectiveness of prioritized telephonic CM within a large, commercially insured population. This is the second study to assess a telephonic CM readmission reduction intervention on a large, commercially insured population.9
All study patients had active private health insurance coverage from the same carrier and were eligible for the same carrier’s CM during the study period between July 1, 2010, and December 31, 2010. All study patients had a 3-day or greater length of stay and an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD- 9-CM) major diagnosis of Heart/Circulatory (ICD-9-CM Major Group CD = 12), Lower Respiratory (ICD-9- CM Major Group CD = 11), or Gastrointestinal (ICD-9-CM Major Group CD = 13) at initial discharge. All major diagnoses were considered, but the 3 major diagnosis categories and 3-day or more length of stay were the inclusion criteria because a previous internal retrospective study revealed those characteristics were most associated with 30-day and 60-day readmissions.10 Discharges were from all 50 states except for Texas and California because those states had existing readmission reduction pilots during the study.
The intervention group was labeled as the prioritized group. The prioritized group received 2 attempted postdischarge phone calls by a case manger within 24 hours of discharge notification. If the prioritized patient was not contacted within 24 hours, then a second attempted call occurred the following day. All attempted calls to the prioritized group were made in descending health risk order (as defined by the Episodic Risk Group [ERG] score) so that outreach was first administered to patients with the greatest likelihood of readmission due to poorer health status. The control group was labeled as the unprioritized group. The unprioritized group received an attempted call by a case manager 3 days after discharge and the unprioritized group calls were not made in any health risk order. If the unprioritized patient was not contacted on day 3, then a second attempted call occurred the following day. If after multiple calls the patient was still not contacted, an Unable to Contact letter was mailed to the patient stating the contact attempts and to return a call.
We created a Call List Flow Chart using 3 days of discharges (ie, 1/3/11 to 1/15/11) to demonstrate the prioritization logic overtime for patients contacted and not contacted (eAppendix A, available at www.ajmc.com). The prioritized patients represent the first half of the call order list receiving initial phone calls in descending risk order within 1 day of discharge, whereas the unprioritized patients represent the second half of the call order list receiving their initial phone call 3 days after discharge and these calls were made in no risk order. One long list was created daily to make the CMs unaware of patient assignment. To further preserve the blinding, 2 CMs for telephonic outreach were scheduled from 8 AM to 5 PM and then 2 different CMs between 5 PM and 9 PM. A total of 86 designated case managers supported the pilot.
The same post-discharge telephonic script (eAppendix B) was provided to the comparison groups. The script’s questions focused on 3 areas: (1) post-discharge understanding of medication, (2) post-discharge care management orders, and (3) scheduling of follow-up visits. An engaged conversation was defined as CM receiving a response to every question. CM phone calls resulting in an unanswered call, voice mail, dropped call, or partial completion of the scripted questions were classified as not engaged. We assumed most participants had an active phone number since the field is a part of annual enrollment information.
Stratified randomization design created comparable prioritized and unprioritized groups at baseline. A daily list of eligible patients discharged from the previous day was produced by utilization managers and called a Census Discharge File. The Census Discharge File contained discharge-related information (ie, dates, diagnosis, procedure codes, etc) describing the inpatient episode and we confirmed that Census Discharge File information was consistent with data in paid claims 90 days later. Our stratified randomization logic assigned study subjects into 1 of several strata defined by a combination of the following variables: gender, ERG score, major ICD-9-CM diagnostic group, count of prior 12-month admissions, health insurance plan type, and hospital facility. Within each stratum, patients were randomly assigned to either the prioritized or unprioritized group. Under this randomization protocol we prospectively matched patients one-to-one on gender, risk score, major diagnostic group, recent admissions history, plan type, and discharge hospital. The matching variables were selected because previous studies showed they were strong predictors of readmissions.5,11-13
The primary outcome was the percent of unique urgent emergent readmissions at 30 days and 60 days. The secondary outcome was readmission rates per 1000. Urgent emergent readmissions were defined as all-cause unscheduled admissions following the initial discharge. We explored measuring clinically related readmissions, but decided against it because secondary claims did not provide the level of clinical detail needed for appropriate measurement, such as chart abstraction information. All outcomes are derived from insurance claims data and CM utilization data including facility, professional, pharmaceutical, and CM call activity.
A group proportions power calculation determined needed sample size. We hypothesized that the intervention would yield a 2% to 3% difference in readmissions between the comparison groups based on the results of a prior internal retrospective analysis.10 Using the power calculation inputs of power 0.80 and a 2-sided P value of .05, the estimated sample size was 3988. We aimed for a cohort of 4786 to account for the 15% to 17% of patients likely to be excluded because of having no comparable match, inactive coverage, or mortality. Baseline demographics and readmission outcomes were analyzed by χ2 tests for discrete variables, 1-way analysis of variance for normally distributed continuous variables, and the Wilcoxon and Mann-Whitney test for skewed distribution of continuous variables. For all statistical analyses, alpha was set to 0.05 and SAS software version 9.1 was used (SAS Institute Inc, Cary, North Carolina). All reported readmission results used intent-to-treat estimation.
Post Hoc Assessments
Three retrospective post hoc assessments further assessed the relationship between the intervention and outcomes. Each post hoc assessment used one-to-one retrospective matched case control without replacement to control for bias due to non-random assignment of the newly created post hoc comparison groups formed from within the original data set. The groups were retrospectively matched on key demographics and were statistically indistinguishable (P >.05) at baseline by age, gender, ERG score, major diagnosis at initial discharge, 12-month admission history, and plan type.
PDF is available on the last page.