Costs and Outcomes in Value-Based Care

Fewer than half of practicing physicians in the United States owned their medical practices in 2016, marking the first time that the majority of physicians were not practice owners (frequently referred to as “in private practice”), according to an American Medical Association study released in 2017. The percentage of physicians who were in hospital-owned practices or employed directly by hospitals hit 32.8 percent, up from 29 percent in 2012.(1) Many factors contributed to this shift, including the market’s movement toward value-based care. A 2016 survey of 465 payers and hospitals showed that 58 percent were moving forward with incorporating value-based care reimbursement protocols into their practices.(2)

Physicians are key stakeholders in the success of value-based care. From providing front-line care to filling leadership positions, physicians are responsible for 75 to 85 percent of all quality and cost decisions,(3) including the selection of drugs and devices used to treat patients. One effective strategy for optimizing this decision-making is transparency regarding cost data and clinical outcomes. Health system leaders, including C-level physician and administrative executives, need to understand the drivers of physician decision-making and present evidence that emphasizes clinical as well as financial outcomes.

The importance of understanding the interdependencies between outcomes and cost was a goal of Physician Perceptions and Practices 2018, a study from Lumere, a firm focused on optimizing health care in the areas of clinical variation and costs. The study gathered responses from physicians about the influences of drug and device preferences and use across various alignment models.

P3 2018 sought to uncover the factors that affect physician decision-making — thereby providing key insights that can help organizations adopt strategically effective alignment models. This deeper understanding of physician drivers is essential as health care spending continues to increase. Average annual inpatient drug spending increased by 23.4 percent between 2013 and 2015, while inpatient drug spending increased on a per-admission basis by 38.7 percent during the same period, according to a report issued by the American Hospital Association and the Federation of American Hospitals.(4) Similarly, data from the financial services firm Moody’s indicate that inpatient drug spending per admission grew 24.1 percent in 2014 and 11.8 percent in 2018. Providers also are spending $175 billion annually on medical devices, according to research from Kalorama Information.(5) What’s more, the worldwide medical device market is expected to grow at a compound annual rate of 4.1 percent through 2020, pushing total industry sales to $477.5 billion, according to a report from EvaluateMedTech.(6) These increased costs have not resulted in improved outcomes for patients in the United States.(7)

Methodology

Purpose: The goal of P3 2018 was to gather insight into physician decision-making, related to improving quality and reducing costs necessary to succeed in a value-based marketplace.

Participants: The study sample of 276 physicians was drawn from the social media network Sermo’s database of U.S. physicians, which includes 40 percent of the U.S. physician population and closely mirrors the profile of all U.S. physicians. Specialties that use medical devices and inpatient drug treatment — general surgeons, orthopedic surgeons, neurological surgeons, cardiothoracic surgeons, interventional radiologists, interventional cardiologists, hospitalists and intensivists — were included in the survey. Sample size was determined by availability of Sermo’s physician panel, selected to optimize response rates and obtain at least 20 responses from each specialty. The response rate for each specialty was 50 to 70 percent. The average respondent had spent 14 years in practice, including residency and fellowship. All respondents were required to have completed their specialty residency training.

Survey Design: The 19-question Likert-style survey was designed to gather insights on physicians’ perceptions of clinical variation, quality improvement strategies, access to clinical outcomes and cost data, involvement in hospital drug and device decision-making, and drivers of drug and device preference. All physicians responded to a set of demographic questions, including gender, years in practice, clinical specialty, participation in alternative payment methods and hospital committee experience. The survey then diverged between sets of surgical specialist-specific questions and medical specialist-specific questions about devices and drugs. In addition to gathering data from employed and independent physicians, the study considered the following health delivery models:

• Accountable care organizations: A group of providers who agree to share responsibility for the quality, cost and coordination of care for a defined patient population.

• Bundled-payment: The payment to health care providers and hospitals for cost over a clinically defined episode of care.

• Co-management: Hospitals and physicians working together to jointly manage the care delivered within a specific service line, with the goal of improving quality and reducing costs.

Survey Analysis: Survey respondents were categorized into four groups, based on experience: less than nine years, nine-12 years, 13-20 years and more than 20 years. Because experience was right-skewed, quartiles were used to establish the range for each group and ensure an even distribution among the groups. Statistical analysis of the groups was performed by Sermo, including pairwise comparisons, correlation coefficients and regression analysis.

Results

Influence 1: Alignment Models

Contrary to the prevailing beliefs of many health system administrators, employed physicians are not necessarily more likely to make decisions based on hospital cost and preference compared to independent physicians. Specifically, when surgical specialists were asked what influences their device selection, employed physicians were equally likely to report that device cost would be “very” or “extremely” influential compared to independent physicians (employed, 48 percent; independent, 48 percent; p-value equals 1). Further, independent physicians might be more likely to report that a health system’s vendor preference would be influential (employed, 33 percent; independent, 39 percent; p-value equals 0.39), although these results are not statistically significant (see Figure 1). This trend suggests employing physicians will not help reduce costs. On further analysis, this might be because employed physicians do not necessarily have a direct financial incentive to save money for the health system.

Figure 1. Impact of alignment models

To understand whether alignment models designed to incentivize cost and quality influence physicians to make cost-effective decisions, we evaluated survey responses from physicians who participate in ACOs, bundled-payment and co-management agreements. Compared to independent physicians who were not part of any alignment model, physicians participating in ACOs, bundled-payment and co-management agreements were slightly less likely to be influenced by device cost when the health system saves (aligned, 46 percent; not aligned, 52 percent; p-value of 0.57), but more likely to be influenced by device cost when they share in savings resulting from the product switch (aligned, 55 percent; not aligned, 46 percent; p-value of 0.40). These results suggest that shared savings might have only a weak effect on physician decision-making (see Figure 2).

Figure 2. Impact of alignment models

Influence 2: Cost

P3 2018 found the vast majority (91 percent) of surgical and medical specialists believe that increasing physician access to cost data would positively affect care quality, but only 40 percent of physicians report their health systems are working to increase physician access to cost data.

In practice, physician incorporation of cost data into clinical decision-making is more nuanced. P3 2018 found experience — both in clinical practice and in evaluating cost data — is directly correlated with the influence of cost data (see Figure 3). This was true regardless of whether the cost is absorbed by the patient (correlation coefficient of 0.13), the physician’s practice (correlation coefficient of 0.22) or the health system (correlation coefficient of 0.12), suggesting that experience with and exposure to cost data might influence physician decision-making. Regression analysis suggests that practice cost and patient cost are significant (beta coefficients of 1.35 percent and 1.07 percent, respectively) and cost to hospitals is not (beta coefficient of 0.46 percent).

Figure 3. Impact of cost data

Further analysis (see Figure 4) reveals physicians involved in committees that make device and technology decisions for their health systems are more likely to report data describing cost to their practice is “extremely” or “very” influential to their clinical decision-making, compared to physicians without committee experience (technology/device committee experience, 62 percent; no technology/device committee experience, 45 percent; p-value of 0.07). Similarly, although not statistically significant, physicians involved in pharmacy and therapeutics committees were slightly more likely to report data describing cost is “extremely” or “very” influential to their clinical decision-making, compared to physicians without P&T experience (P&T experience, 58 percent; no P&T experience, 48 percent; p-value of 0.52).

Figure 4. Impact of physician practice cost data

Based on these results, physicians with more exposure to cost data and variation reduction efforts (via years of experience or participation in those efforts) generally find cost data more influential in their clinical decision-making. Providing physicians with in-depth exposure to cost data might be an effective strategy for helping physicians make cost-effective decisions.

Influence 3: Variation

Reducing clinical variation generally has been considered one of the more comprehensive strategies for optimizing cost and quality of care. However, variation reduction is not as commonly pursued as other strategies. Only 57 percent of respondents said their hospitals or health systems are reducing variation in practice patterns, 53 percent said their organizations are reducing variation in the drugs selected for a clinically equivalent patient population, and 52 percent said their hospitals or health systems are reducing variation in the devices selected for a clinically equivalent patient population. This is compared to 82 percent of institutions that are implementing clinical practice guidelines, 75 percent that are improving care coordination and sharing of information among providers, and 64 percent that are increasing access to quality data. However, medical and surgical specialists both reported variation data will help improve the quality of care when incentivized by value-based payment models.

Further clarifying the importance of variation data, the majority (86 percent) of surgical and medical specialists reported that increasing physician access to clinical practice variation data would positively impact care quality, 11 percent reported it would not have a negative or positive impact, and 3 percent reported it would be detrimental.

Again, experience in practice increased variation data’s influence on clinical decision-making, with 67 percent of physicians with more than 20 years’ experience describing variation data as “very” or “extremely” influential, compared to only 32 percent of physicians with less than nine years’ experience (see Figure 5). There is some strength to this relationship between utility of variation data and experience, with a correlation coefficient of 0.14 and a beta coefficient of 0.0083. We might conclude that for every 10 years in practice, the influence of variation data increases 8 percent.

Figure 5. Impact of variation reduction initiatives

Encouragingly, physicians who have experience with variation reduction efforts also report that variation data is highly influential to clinical decision-making, especially regarding variation in drug and device selection (see Figure 6).

Figure 6. Impact of variation reduction initiatives on quality improvement

Nearly a quarter of respondents did not know if their hospitals or health systems were implementing variation reduction strategies. This lack of exposure might contribute to the ineffectiveness of variation reduction efforts. Building physician support might require providing physicians with information on the relationship between reducing clinical variation and improving outcomes. In addition, as variation reduction becomes more common, it could become a more-influential decision-making factor.

Discussion

P3 2018 contains several demonstrable findings, namely that:

• Physician employment alone has little bearing on truly “aligning” physician behavior.

• Years and depth of practice experience are correlated with physician or health system alignment goals of improving value-based care.

• Data transparency is a strong driver for positively influencing physician behavior.

Based on our study, a strong first step for health care organizations that want to optimize physician engagement is to critically assess how data is systematically shared with physicians. Health care leaders should then develop a strategy regarding the appropriate amount and type of data to share as part of an ongoing educational program. Quality, outcome and cost data should be benchmarked on contemporary, evidence-based guidelines whenever possible and regularly shared with physicians in a consistent and usable format. Health systems will gain an early advantage by openly and transparently engaging with physicians and fostering a culture that use data and analytics to improve the value of care.

While the move toward value-based care and new payment models has been in motion for some time, greater transparency of data will be required in many settings to be successful in engaging physicians as advanced risk-sharing payment models evolve.

Further supporting these recommendations, survey respondents expressed a clear willingness to incorporate reliable data in their decision-making (see Figure 7). In addition, physicians expressed the most support for improving care coordination and sharing information among providers. Health system leaders should acknowledge physicians’ recognition of the relationship between continuity of care and quality of care — and leverage this understanding to improve value to the patients they serve.

Figure 7. Impact of involvement in an evidence-based approach

The P3 2018 findings indicate cost and variation transparency might help health care organizations reduce cost and variation, which is best accomplished by providing physicians with reliable clinical evidence and outcomes data to inform selection of drugs and devices. These findings are increasingly important as the health care industry continues to feel pressure to spend more on drugs and devices, even though these increased costs do not necessarily correlate to improved value. For example, a study published in The BMJ showed that five widely used and marketed advances in hip and knee replacement technology did not result in improved clinical outcomes.

Finally, survey respondents’ support for data transparency and adoption of clinical practice guidelines suggests opportunities exist to reduce variation in practice patterns and cost. This knowledge of physician’s attitudes and preferences — and the reasons behind these heuristics — is valuable information that can help hospital administrators and physician leaders better partner with physicians and assist them in making the best drug, device and clinical decisions.

Acknowledgements

With contributions from Manuel Astorga, survey data operations manager for Sermo, a global social network for physicians.

References

1. Kane C. American Medical Association. Policy Research Perspectives. Updated Data on Physician Practice Arrangements: Physician Ownership Drops Below 50 Percent. May 2017. https://www.ama-assn.org/sites/default/files/media-browser/public/health-policy/PRP-2016-physician-benchmark-survey.pdf.

2. Press Release. Value-Based Payment Hits the Tipping Point. https://www.businesswire.com/news/home/20160613005067/en/Value-Based-Payment-Hits-Tipping-Point.

3. Oshiro B. 6 Proven strategies for engaging physicians and 4 ways to fail. Health Catalyst.https://www.healthcatalyst.com/proven-physician-engagement-strategies.

4. American Hospital Association. Drug Costs: Issues and Challenges. https://www.aha.org/system/files/2018-01/aha-fah-rx-report.pdf.

5. Kalorama Information. The Global Market For Medical Devices, 8th Ed. Oct. 6, 2017. https://www.kaloramainformation.com/Global-Medical-Devices-11198459/.

6. Evaluate. Evaluate MedTech World Preview 2015, Outlook to 2020.http://www.evaluategroup.com/public/reports/EvaluateMedTech-World-Preview-2015.aspx.

7. OECD Health Statistics. https://www.oecd-ilibrary.org/social-issues-migration-health/data/oecd-health-statisticshealth-data-en.

Additional Resources

Camacho J, Medina Ch AM, et al. Comparing a mobile decision support system versus the use of printed materials for the implementation of an evidence-based recommendation: protocol for a qualitative evaluation. JMIR Res Protoc. 2018 Apr 13;7(4):e105

Fralicx R. Strange bed(side) fellows physician-finance collaboration. Healthcare Finance Management. 2012 Jul;66(7):90-4, 96.

Goldsmith J, Hunter A, Strauss A. Do most hospitals benefit from directly employing physicians? Harvard Business Review. 2018 May 29. https://hbr.org/2018/05/do-most-hospitals-benefit-from-directly-employing-physicians.

Michaels, D. Implementing A New Compensation Plan: 5 Ways to Get Physician Buy-In. Hayes Management. http://meetings.hayesmanagement.com/blog/implementing-a-new-compensation-plan-5-ways-to-get-physician-buy-in .

Guzman MJ, Gitelis ME, et al. A model of cost reduction and standardization: improved cost savings while maintaining the quality of care. Dis Colon Rectum. 2015 Nov;58(11):1104-7.

Morris M, Abrams K, et al. Practicing Value-Based Care: What Do Doctors Need? Deloitte, 2016. https://www2.deloitte.com/content/dam/insights/us/articles/3140_Practicing-value-based-care/DUP_Practicing-value-based-care.pdf.

Zygourakis CC, Valencia V, et al. Association between surgeon scorecard use and operating room costs. JAMA Surg. 2017;1;152(3):284-91. doi:10.1001/jamasurg.2016.467.