Currently Viewing:
The American Journal of Managed Care January 2018
Measuring Overuse With Electronic Health Records Data
Thomas Isaac, MD, MBA, MPH; Meredith B. Rosenthal, PhD; Carrie H. Colla, PhD; Nancy E. Morden, MD, MPH; Alexander J. Mainor, JD, MPH; Zhonghe Li, MS; Kevin H. Nguyen, MS; Elizabeth A. Kinsella, BA; and Thomas D. Sequist, MD, MPH
The Health Information Technology Special Issue: Has IT Become a Mandatory Part of Health and Healthcare?
Jacob Reider, MD
Bridging the Digital Divide: Mobile Access to Personal Health Records Among Patients With Diabetes
Ilana Graetz, PhD; Jie Huang, PhD; Richard J. Brand, PhD; John Hsu, MD, MBA, MSCE; Cyrus K. Yamin, MD; and Mary E. Reed, DrPH
Electronic Health Record "Super-Users" and "Under-Users" in Ambulatory Care Practices
Juliet Rumball-Smith, MBChB, PhD; Paul Shekelle, MD, PhD; and Cheryl L. Damberg, PhD
Electronic Sharing of Diagnostic Information and Patient Outcomes
Darwyyn Deyo, PhD; Amir Khaliq, PhD; David Mitchell, PhD; and Danny R. Hughes, PhD
Hospital Participation in Meaningful Use and Racial Disparities in Readmissions
Mark Aaron Unruh, PhD; Hye-Young Jung, PhD; Rainu Kaushal, MD, MPH; and Joshua R. Vest, PhD, MPH
Currently Reading
A Cost-Effectiveness Analysis of Cardiology eConsults for Medicaid Patients
Daren Anderson, MD; Victor Villagra, MD; Emil N. Coman, PhD; Ianita Zlateva, MPH; Alex Hutchinson, MBA; Jose Villagra, BS; and J. Nwando Olayiwola, MD, MPH
Racial/Ethnic Variation in Devices Used to Access Patient Portals
Eva Chang, PhD, MPH; Katherine Blondon, MD, PhD; Courtney R. Lyles, PhD; Luesa Jordan, BA; and James D. Ralston, MD, MPH
Hospitalized Patients' and Family Members' Preferences for Real-Time, Transparent Access to Their Hospital Records
Michael J. Waxman, MD, MPH; Kurt Lozier, MBA; Lana Vasiljevic, MS; Kira Novakofski, PhD; James Desemone, MD; John O'Kane, RRT-NPS, MBA; Elizabeth M. Dufort, MD; David Wood, MBA; Ashar Ata, MBBS, PhD; Louis Filhour, PhD, RN; & Richard J. Blinkhorn Jr, MD

A Cost-Effectiveness Analysis of Cardiology eConsults for Medicaid Patients

Daren Anderson, MD; Victor Villagra, MD; Emil N. Coman, PhD; Ianita Zlateva, MPH; Alex Hutchinson, MBA; Jose Villagra, BS; and J. Nwando Olayiwola, MD, MPH
A randomized trial of eConsults for cardiology referrals from primary care resulted in significant reductions in total cost of care compared with traditional face-to-face consultations.
ABSTRACT

Objectives: To evaluate the cost-effectiveness of electronic consultations (eConsults) for cardiology compared with traditional face-to-face consults.

Study Design: Cost-effectiveness analysis for a subset of Medicaid-insured patients in a cluster-randomized trial of eConsults versus the traditional face-to-face consultation process in a statewide federally qualified health center.

Methods: A total of 369 Medicaid patients were referred for cardiology consultations by primary care providers who were randomly assigned to use either eConsults or their usual face-to-face referral process. Primary care providers in the eConsult arm transmitted consults to cardiologists using a secure peer-to-peer communication platform in an electronic health record. Intention-to-treat analysis was used to assess the total cost of care and cost across 7 categories: inpatient, outpatient, emergency department, pharmacy, labs, cardiac procedures, and “all other.” Costs are from the payer’s perspective.

Results: Six months after the cardiology consult, patients in the eConsult group had significantly lower mean unadjusted total costs by $655 per patient, or lower mean costs by $466 per patient when adjusted for non-normality, compared with those in the face-to-face arm. The eConsult group had a significantly lower cost by $81 per patient in the outpatient cardiac procedures category.

Conclusions: These findings suggest that eConsults are associated with total cost savings to payers due principally to reductions in the cost of cardiac outpatient procedures.

Am J Manag Care. 2018;24(1):e9-e16
Takeaway Points
Electronic consultations (eConsults) improve access, timeliness, and coordination of care compared with traditional face-to-face consultations. Findings from this study suggest that the use of eConsults is associated with cost savings to payers due principally to reductions in the cost of cardiac outpatient procedures.
  • The implications of cost savings demonstrated in this study are important for state Medicaid agencies and other health systems seeking new ways to improve access and quality while reducing cost.
  • Policy changes that support the use of eConsults could result in significant savings to the Medicaid program in a relatively short time frame. 
Many initiatives aimed at transforming primary care have concentrated on the development of patient-centered medical homes, with emphasis on elements including the adoption of electronic health records (EHRs), multidisciplinary team-based care, and care coordination. Fewer efforts have been directed at improving the interface between primary care providers (PCPs) and specialists in the outpatient setting.1-3 This gap is notable given the significant clinical importance and financial impact of the PCP–specialist relationship. Outpatient specialty visits represent a disproportionate source of year-over-year increases in healthcare expenditures,4,5 with research suggesting that a typical PCP interacts with more than 200 specialists in a year.6 Such financial considerations are increasingly important as payment reform gains momentum across the country and stimulates experimentation with novel reimbursement arrangements. Additionally, the proliferation and adoption of new technologies, including EHRs and secure health information exchanges, are creating fertile conditions for improving the interface between specialists and PCPs. 

Electronic consultations (eConsults) are non–face-to-face (F2F) consultations between a PCP and a specialist that utilize secure messaging to exchange information. Unlike electronic referral systems that link primary care practices with specialty providers for F2F appointment triage, eConsults provide a virtual consultation by the specialist after clinical information sent by the PCP is reviewed and returned with recommendations, which potentially eliminates the need for the patient to be seen in person by the specialist. Health systems that implemented eConsults have improved specialty access, reduced wait times,7 and decreased F2F consultations between 9% and 51% depending on setting and specialty.8-14 However, few studies have evaluated the effects of PCP access to a secure eConsult platform on total healthcare expenditures. Findings using retrospective data from an eConsult program in Canada suggest the potential for cost savings,15,16 but these studies were not randomized and did not evaluate the impact on total cost of care. The reduction in F2F visits with specialists is a potential source of cost savings to payers, but these savings could be offset by an increase in primary care costs and the cost of administering an eConsult program. We recently published results of a cluster-randomized controlled trial of eConsults for cardiology in a statewide federally qualified health center (FQHC) in Connecticut14 that demonstrate significant improvements in access and timeliness of care with a reduction in cardiology utilization. In this article, we report the impact of the intervention on cost for the subset of Medicaid-insured patients in this trial.

METHODS

Setting

Community Health Center, Inc, (CHCI) is a statewide multisite FQHC providing comprehensive primary medical, behavioral, and dental care to medically underserved patients in Connecticut. CHCI delivers care in 13 primary care clinics as well as in numerous school-based and homeless shelter–based facilities. All sites use an integrated EHR. Patients receive primary medical care from internists, family physicians, pediatricians, nurse practitioners, and physician assistants. Most of CHCI’s practice sites refer to hospitals and specialists within their neighboring communities or to large regional academic medical centers. During the study, more than 60% of CHCI’s patients were racial/ethnic minorities, more than 90% had incomes at or below 200% of the federal poverty level, more than 60% had state Medicaid insurance, and almost 25% were uninsured.

Study Design

Complete details of the design and methods of the trial have been published.14 Briefly, the intervention period for the eConsult study was between August 1, 2012, and June 30, 2013, and involved 590 patients and 36 providers from CHCI and 3 cardiologists from the University of Connecticut Health Center (UCHC). All consenting PCPs were assigned to the intervention (eConsult) or control (F2F) arm using 1:1 blocked randomization at the level of the PCP. No other parameters were used. There were no significant differences in site of practice between the intervention and control sites. All providers at all practices accepted all patients regardless of insurance status.

Intervention providers used eConsults for all nonurgent cardiology referrals except for patients who had an established relationship with a cardiologist. Determination of urgency was at the discretion of the PCP. The eConsult option was a function embedded in the EHR that allowed direct electronic communication between the PCP and the cardiologist. The eConsult included a specific question and relevant documentation, such as a brief clinical history, electrocardiograms, medication lists, laboratory and procedure results, and progress notes. A referral coordinator managed the eConsult process. The participating cardiologist received an email notification each time an eConsult was submitted, retrieved the eConsult from a secure Web portal, and responded within 2 business days. Their responses generally provided answers to PCPs’ questions and included other relevant suggestions, such as additional laboratory readings/tests or therapeutic trials prior to a subsequent consult, or occasionally a recommendation for a F2F visit. When a F2F consultation was recommended, providers and patients were free to choose any cardiologist accepting FQHC referrals in the service area. Providers in the control group sent all cardiology consults via the traditional F2F referral process at CHCI (Figure). The institutional review board of CHCI approved the study.

Data Sources

The economic analysis used demographic information for participating PCPs and their patients from CHCI’s practice management system and Medicaid paid claims data between August 8, 2011, and February 21, 2014.

Statistical Analysis

Three types of analysis were conducted: 1) an intention-to-treat (ITT) analysis, 2) an analysis of actual treatment (AT) received, and 3) a sensitivity analysis. 

In the ITT analysis (Group B vs C in the Figure), all claims from patients in the PCP intervention and control arms were counted in their respective groups regardless of provider’s or patient’s adherence to their assigned consultation arm. 

In the AT analysis, patients were grouped based on actual consultation choice (eConsult vs F2F), regardless of the provider’s assigned group. This second analysis regrouped claims of patients of intervention PCPs who were reassigned to a F2F consult as per the study protocol (B+E vs F in the Figure). This analysis presents the postrandomization (“real-world”) provider referral behavior. 

The sensitivity analysis used 3 hypothetical fee combinations. All combinations were tested for the ITT and AT scenarios. In addition to the $25 eConsult fee charged for this study, we used $185 per visit for F2F visits and $45 per visit for eConsults. The latter 2 reimbursement rates correspond to the average commercial reimbursement rate for a 30-minute new patient F2F office consultation in the same zip code as the UCHC and a cost-based estimate for eConsults, respectively.17,18 

Cost items were segregated into the categories shown in Table 1. Baseline costs were established by evaluating all claims for 180 days preceding the cardiology consult request. Cost analysis for the intervention period was based on claims inclusive of the date of the referral and the following 180 days. All claims included a 3-month lag. Extreme costs were not truncated.

All cost analyses were performed from the payer’s (Connecticut Medicaid) perspective. Transportation costs paid for by Medicaid for F2F visits were not included. At the time of the study, Medicaid did not reimburse for eConsults and therefore payment was not reflected in the claims extracts. All analyses included a $25 fee for each eConsult visit paid to the cardiologist by CHCI. Cardiology F2F new visits (Current Procedural Technology code, 99243) were reimbursed by Medicaid at their customary rate of $66. Any costs borne by PCPs (eg, additional time spent creating and reviewing eConsults), specialists (eg, lost revenue from “no-shows”), or patients (eg, co-pays, unpaid time off work, or out-of-pocket transportation costs) were not included. 

Healthcare costs are typically not normally distributed (ie, they are skewed),19 resulting in the distributions of repeated cost variables being “pulled up” toward a higher mean by a few extreme scores. Several statistical paths were followed to ensure that comparisons of changes in costs between F2F and eConsult patients yielded robust results. 

Baseline and intervention costs were assessed across 7 categories (inpatient, emergency department, outpatient, pharmacy, labs, cardiac procedures, and all other) for departure from normality. Then, non-normal cost changes were modeled using Mplus version 7.4 software (Muthén & Muthén; Los Angeles, California).20,21 Its skew-t estimation method allows for direct comparisons of means without the need to truncate scores, by estimating 2 parameters beyond mean and variance, namely the skewness and t degrees of freedom for extreme scores (to model “thick-tailed” distributions).22 

Patient demographic characteristics and raw baseline costs were first evaluated for baseline equivalency. This was followed by analyses of differences between the non-normality–adjusted means of the cost changes (ie, change scores adjusted for baseline values).23,24 All results are reported as the test of differences in changes between cost categories from the baseline to intervention periods for the total cost. Amounts paid in 2013, 2014, and 2015 were converted to 2016 dollars. All claims categories (ie, cardiac and noncardiac) were included in the analysis.

RESULTS

Thirty-six PCPs participated in the trial; 19 were randomly assigned to the control group and 17 to the intervention group. Characteristics of the PCPs in both groups were balanced, with no statistically significant differences in age, clinical experience, gender, race/ethnicity, or primary care specialty (Table 2). 

During the study period, these participating PCPs initiated 590 adult cardiology consults. Of those, 369 patients had Medicaid insurance continuously for the duration of the study and were pooled for this comparative cost analysis. 

 
Copyright AJMC 2006-2018 Clinical Care Targeted Communications Group, LLC. All Rights Reserved.
x
Welcome the the new and improved AJMC.com, the premier managed market network. Tell us about yourself so that we can serve you better.
Sign Up
×

Sign In

Not a member? Sign up now!