Actual Case Study

How An Efficient, Quality Practitioner Network Reduced Health Care Expenditures--Documented Savings

The Pitney Bowes Experiment is detailed in an article entitled, "Pitney Bowes: Using Comprehensive Cost Information to Build Provider Networks" (Cave et al 1995, Benefits Quarterly 11(3):6-23). Below, we summarize key excerpts from this article. The analysis was performed in 1993, and the efficient network was implemented in 1994.

Plan Offerings: Before and After the Experiment

In 1993 (the year before program implementation), Pitney Bowes Fairfield County employees were enrolled in three different health plans: MetLife (an indemnity program with a PPO offering), Physician Health Services (PHS, an HMO with no out-of-network benefit), and ConnectiCare (an HMO with no out-of-network benefit). Employee distribution in these plans was as follows: 75% in MetLife, 20% in PHS, and 5% in ConnectiCare.

The MetLife indemnity/PPO plan was designed to direct employees to PPO providers, although the channeling incentive was not strong (approximately a 10% difference in coinsurance between in-network and out-of-network utilization). About 70% of employee expenditures were incurred through PPO providers. The MetLife indemnity/PPO plan had inpatient utilization review (UR) and large case management programs. Approximately 50% of all Fairfield County practitioners accessed the MetLife PPO network contract and fee discounts. Practitioners were paid on a discounted fee-for-service (FFS) basis, and hospitals were paid on a per diem basis.

The new health plan program became effective January 1, 1994. Pitney Bowes now offered only two plans, both administered by PHS: PHS point-of-service (POS) plan, and PHS exclusive provider organization (EPO) plan. MetLife and ConnectiCare were no longer offered in 1994.

The POS plan had a $10 office visit copay, 80%/20% out-of-network coinsurance, and no gatekeeper mechanism. The EPO had a $10 copay, no out-of-network benefit, and no gatekeeper mechanism. Both the POS and EPO plans had inpatient UR and large case management programs.

PHS's Fairfield County network was comprised of about 30% of the practitioners providing care for MetLife's indemnity/PPO book-of-business in Fairfield County. There were some practitioners in PHS's network that were not under MetLife network contracts. The PHS contracted network size was about 1,000 practitioners. Practitioners were paid on a discounted FFS basis.

The Initial Health Plan/Practitioner Efficiency Study

In 1993, a study was performed to compare episode practice patterns for practitioners in MetLife's indemnity/PPO plan versus those under contract in PHS's HMO. The study included two populations of Fairfield County Pitney Bowes' employees (and their dependents). The first population consisted of Pitney Bowes employees and their dependents enrolled in the MetLife indemnity/PPO plan. Approximately 75% of employees and dependents were enrolled in the MetLife indemnity/PPO plan at the time of the study. The second population comprised members enrolled in a large HMO-PHS. The membership base for this population consisted not just of Pitney Bowes employees, but PHS's book of business, derived from over 35 employers in a wide variety of industries.

Details of the episode methodology are provided in the 1995 Benefits Quarterly article. For primary care physicians (PCPs), results of the analysis showed that average longitudinal episodes-of-care charges were 34% lower (P<0.01) for the PHS (the HMO) PCPs as compared to the same medical condition and SOI level episodes that were treated by MetLife indemnity/PPO PCPs. For specialists, results showed that average episode-of-care charges were 25% lower (P<0.01) for the PHS specialists as compared to the MetLife indemnity/PPO specialists.

Practitioner Composition of PHS POS and EPO Networks

Study results showed that PHS contracted practitioners delivered care significantly more cost effectively than MetLife indemnity/PPO practitioners. Consequently, PHS decided to move all employees to PHS's practitioner network. Pitney Bowes implemented two health plans, using PHS to administer the plans:

• PHS POS plan. All practitioners under contract with PHS remained in the PHS network. No PCPs or specialists were eliminated. Under this plan, about 1,000 practitioners provided care to the POS enrolled Fairfield County employees (and their dependents). The POS plan had a $10 copay, 80%/20% out-of-network coinsurance, and no gatekeeper mechanism.
  
  
• PHS EPO plan. This network was a subset of the PHS POS plan. The top 100 inefficient practitioners (i.e., used significantly more resources to treat the same medical condition episodes) were eliminated from the EPO network. This exclusion included PCPs and practitioners from almost every specialty type. The EPO plan had a $10 copay, no out-of-network benefit, and no gatekeeper mechanism.

Monthly employee premium contributions were reduced for employees that enrolled in the EPO plan. Consequently, there was good EPO plan enrollment in the 1994 open enrollment period--about 45% of all Fairfield County employees.

Program Savings

The program became effective on January 1, 1994. All employees were moved from the MetLife indemnity/PPO plan and ConnectiCare HMO to the PHS network of providers. After open enrollment, 55% of employees enrolled in the PHS-administered POS plan, and the remaining 45% of employees enrolled in the PHS-administered EPO plan.

"In the first year of implementation (1994), the company saved 19.3% from expected cost increases (excluding the expense of administering this program--which was estimated to be small)." Within Connecticut, where healthcare trend rose an average of 10%, Pitney Bowes experienced a 9.3% decrease in per employee costs--or a 19.3% absolute difference from expected.

• In the pre-implementation year (1993), per employee costs were calculated by adding together the PHS and ConnectiCare HMO total premium costs and the self-insured MetLife indemnity/PPO plan costs. The MetLife administrative services charges were included. This total was divided by the total number of employees electing coverage in one of the three offered health plans.
  
  
• In the post-implementation year (1994), per employee costs were calculated by adding together the PHS POS and EPO costs (no fixed premiums were charged by PHS). The PHS administration services charges were included. This total was divided by the number of employees selecting POS and EPO plan coverage.

Copyright © 2003 Cave Consulting Group, All Rights Reserved