The Commonwealth Fund Health Care Opinion Leaders Survey: Assessing Health Care Experts' Views on Health Care Costs

Country: United States

Survey Organization: Harris Interactive, Inc.

Field Dates: April 7, 2005–April 21, 2005

Sample: Peer-nominated experts in health care policy, finance, and delivery

Sample Size: 289 (survey successfully e-mailed to 1,256 potential respondents; 23% response rate)

Interview Method: Online/E-mail

The Commonwealth Fund Health Care Opinion Leaders Survey was conducted by Harris Interactive on behalf of The Commonwealth Fund, with a broad group of 289 opinion leaders in health policy and innovators in health care delivery and finance. It was the third in a series of six bimonthly surveys designed to highlight leaders' perspectives on the most important and timely health policy issues facing the nation. Respondents were clustered into four main groups: those employed by academic or research organizations; those involved in the delivery of health care services; those employed by businesses or health industry including health insurance and managed care plans; and those working for government or labor/consumer advocacy organizations.

This survey focused on potential ways to lower health care spending, addressing key components of spending such as prices charged, utilization levels, and insurance overhead. Potential respondents for the surveys in this series were identified through a two-step process involving 1) a "nomination" survey with a core group of experts in multiple fields to nominate additional leaders both within and outside their areas of expertise and 2) a review of published lists and directories of recognized health experts. The detailed methodology is provided in the Appendix.

Summary
Health care spending in the United States continues to rise at a level far outpacing the rate of inflation. Finding effective ways to control these expenditures is a central concern for policymakers, those in the public and private health sectors, and consumers. Health care spending depends on prices charged and utilization levels. We asked respondents to rate the effectiveness of different initiatives and actions to reduce prices and utilization. We also focused on a range of options to lower the costs associated with insurance overhead.

In general, majorities of our panelists consider all of the actions presented to them at least somewhat effective. For controlling prices, there is considerable agreement that some type of a pay-for-performance approach encouraging medical providers to lower costs and improve quality would be most effective. To lowering utilization of health services, panelists believe that better management of high-cost conditions and the use of evidence-based treatment guidelines would be most effective. Finally, to reduce insurance overhead, panelists believe that having private insurance and public programs working together to streamline and standardize their products and processes would be most effective.

Prices: When asked about ways to control prices of health care services, most opinion leaders (57%) rate rewarding more efficient and high-quality medical care providers as extremely or very effective. This action ranks as the most effective way of controlling costs across all health care sectors. Having all payers (private, Medicare, and Medicaid) adopt common payment methods and rates ranks as the second most effective option to control prices, with slightly fewer than half of respondents rating it as extremely or very effective. Other initiatives receive less support, with about one-third of respondents considering them as extremely or very effective. These are: promoting best practices and supporting provider learning collaboratives to improve efficiency and quality; making public information available on comparative quality and total costs of care; and providing feedback with comparative information on total resource consumption and quality to physicians and hospitals.

Utilization: Two options emerge as the most effective ways of reducing unnecessary utilization of health care services, with slightly more than half of respondents rating them as extremely or very effective. With some small variations among the sectors, improving disease management services for patients with high-cost conditions and enhancing primary care case management ranks as the most effective action (56% of all respondents found this to be a highly effect strategy). It is followed closely by using evidence-based medicine guidelines or protocols to determine when a given test or procedure should be done (52% of all respondents found this highly effective). Expanding the use of information technology ranks third, receiving ratings of extremely or very effective from slightly fewer than half of respondents. Fewer respondents believe in the effectiveness of implementing better measures of and reporting on over-utilization and having consumers pay a substantially higher share of their health care costs, with about one-third rating them as extremely or very effective.

Overhead: According to respondents from all sectors, the most effective way to reduce high insurance overhead is to increase collaboration among public programs and private insurers to streamline administrative costs, including standardizing insurance products and processes. About two of five (41%) thought this would be highly effective. However, when presented with all other possible actions, there is considerably more skepticism about their effectiveness and the sectors differ greatly in their opinions as to what would and would not work. Slightly fewer than one-third to about a quarter of respondents view the following actions as extremely or very effective: making health insurance a public utility regulated by states; creating a more competitive market with strong competition among different insurers; and creating a state electronic clearinghouse with consolidated electronic information on enrollees and claims.

The topline results, survey tables, and summary of key findings are attached at right. You may also be interested in commentary by Fund President Karen Davis, or commentaries on reducing health care costs by Helen Darling, president of the National Business Group on Health, and Alan R. Nelson, M.D., special advisor to the CEO of the American College of Physicians and a former president of the American Medical Association.

The goal for this survey is to expand and inform a healthy public and professional debate. As part of that effort to keep the discussion going, we invite you to take the survey yourself.

Methodology
The online survey was conducted by Harris Interactive with 289 opinion leaders in health policy and innovators in health care delivery and finance between April 7, 2005, and April 21, 2005.

The sample for this survey was developed through a two-step process. Initially, The Commonwealth Fund and Harris Interactive jointly identified a number of experts across different industries and professional sectors with a range of perspectives, based on their affiliations and involvement in various organizations and institutions. Harris Interactive then conducted an online survey with these experts asking them to nominate others within and outside their own fields whom they consider to be leaders and innovators in the health care industry. Based on the result of the survey and after careful review by Harris Interactive, The Commonwealth Fund, and a selected group of health care experts, the sample for this poll was created. The final list included 1,314 people.

Harris Interactive sent out individual e-mail invitations containing a password-protected link to the survey to the entire sample. Of the 1,314 email invitations, 58 were returned as undeliverable, resulting in a final sample of 1,256. Steps were taken to attempt to correct the e-mail addresses and locate the individuals, however these efforts were unsuccessful. Harris Interactive determined that the undeliverable e-mail addresses appeared to be randomly distributed among the different sectors and affiliations. Data collection took place between April 7, 2005, and April 21, 2005. A total of three reminders were sent to anyone who had not responded. The response rate was 23 percent. Typically, samples of this size are associated with a sampling error of +/- 6%.