The Current State of Healthcare Quality Measures
Much has been written about the current state of healthcare and the dire implications for all stakeholders should we continue down the current path. Thousands of studies around variance in healthcare abound where the words of Hebert Simon resound for me, “A wealth of information creates a poverty of attention.” In close to 3 decades of work with hundreds of organizations I’ve seen no material progress in systemic change of the scale required to improve the plight of consumers, employers, and providers in achieving and rewarding greater value. Solutions may be within our reach but clearly not within our grasp. Quite maddening. Michael Porter asserts a very simple framework for all participants around patient value. He declares, “Achieving high value for patients must become the overarching goal of healthcare delivery, with value defined as the health outcomes achieved per dollar spent. This goal is what matters for patients and unites the interests of all actors in the system. If value improves, patients, payers, providers, and suppliers can all benefit while the economic sustainability of the healthcare system increases.” It remains to be seen if this is the destination our current course and trajectory delivers. In my opinion, one of the significant barriers inhibiting progress in higher patient value is the lack of knowledge in proper measurement of cost and quality, the two essential components in patient value creation.
If an essential pillar of value is measurement of quality, how is it that there are no clear or consistent approaches which offer a firm foundation? When ProPac moved to prospective payment in 1983 with the implementation of DRGs, the criticism of measuring mortality then was very similar to the current discourse around the current star ratings produced by CMS. The criticism in both instances is based on substantive arguments on the lack of appropriate risk adjustment, and improper aggregation, which falls short of clear, precise, actionable information. My personal experience when sharing information around quality variation to non-provider stakeholders is surprise, bewilderment, and confusion. This knowledge presents more problems to the purchasers and consumers of healthcare making the aim of higher patient value virtually unattainable. The industry has been measuring inpatient quality only where the encounters may be infrequent, but the impact of adverse events can lead to serious harm and much higher cost. The fact is the preponderance of care is delivered in outpatient settings where there is virtually no quantitative data about provider quality performance. This challenge is rapidly compounding as we now are increasingly moving to outpatient settings with no visibility to provider quality. Services delivered in outpatient settings tend to be less costly achieving one objective of greater value. So, do we take a leap of faith with respect to quality assuming all quality is good? Do we accept surrogates and proxies for actual performance? Do we believe the advice of friends, family and physicians with no actual information about the performance of those performing complex, and dangerous, procedures on us or our loved ones?
I would suggest that components of the current scientific framework for assessing provider quality is extensible to outpatient settings and can provide the information to determine whether “high patient value” is being achieved. I believe there are three components required:
1) New Risk-Adjustment Models
New risk models for quality outcomes assessment must be calibrated across delivery settings and have yet to be developed. Severity adjustment, and risk-adjustment, are terms widely used in the context of inpatient care with their origins beginning over 40 years ago. Many current methods are iterations of the same approaches derived in the 80’s still largely limited to outcomes of inpatient care only. Much of this information is not a valid proxy for outpatient provider performance, especially if the procedures are performed primarily in outpatient settings. Equally important, calibrations of current risk models themselves are based on inpatient encounters alone thus not reflecting the characteristics of patients receiving care in outpatient settings. I believe an episodic framework for risk models calibrated to procedures, and medical conditions, regardless of care setting is a vital component of ascertaining quality across all delivery settings. We can rely on the same algorithmic approaches, which have served us well, but consider patient risk much differently. For example, total knee replacements are increasingly being performed in outpatient settings. If we are looking at measuring complications of care, risk adjusted, the propensity for a complication must now be contextualized to both inpatient and outpatient settings together. Indicators can be global measures of complications or specific events such as ED visits, observation encounters, hospitalizations, or readmission’s. Combined with more homogeneous procedure groups, and medical conditions, this new approach provides a unified, comprehensive, assessment of quality for providers intent on improving outcomes as well as purchasers directing care towards higher quality. With this approach providers can be accurately assessed across a much larger range of care delivery spanning minor procedures with virtually no risk, to complex procedures across different delivery settings. The range of quality variance across providers will be compounded significantly as compared to the inpatient assessments currently being made. This huge first step of a patient specific risk adjustment is essential if we aspire to understand the quality delivered across providers and settings.
2) Quality Scoring Approach
The second requirement is a quality scoring approach that renders precise, concise, ratings of providers spanning inpatient and outpatient care. “Don’t confuse me with the facts.” This is essentially the intellectual strategy I see used constantly to dismiss, or avoid, measuring provider quality whether inpatient or outpatient. Would you purchase a car by assessing the cleanliness of the dealership restroom? Would you research restaurants in a city to purchase a new home sight unseen? While these are seemingly hyperbole, much of what is currently used to gauge quality in healthcare uses the same logic. Hospital surveys, reputation, word of mouth, imputing quality from one clinical area, or setting, to another along with wildly incorrect analytical approaches inhibit helpful, and accurate, quality scoring for all stakeholders. Presenting quality performance for specific clinical areas across thousands of hospitals, and hundreds of thousands of physicians, across all settings can be done. Smashing data together and creating arbitrary rules for aggregating observations is not only imprecise, its dangerous! CMS currently delivers hospital ratings that are scaled so that almost 97% of all hospitals are considered “average”. Nothing could be further from the truth. It is irresponsible to suggest that a largely illiterate public should rely on that information to select providers. Would it also make sense to use the CMS all-cause mortality rate for pneumonia to determine where I should have an outpatient colonoscopy? Clearly not. Meaningful quality scoring must use indicators of quality for the procedure, or condition, for which care will be delivered. Helpful quality scoring must convey the relative performance of providers in avoiding adverse events for inpatient and outpatient care.
Quantros uses an approach which measures the levels of statistical significance across all the indicators we can reliably measure for specific areas with the providers responsible for that care. In a sense, we transpose multiple risk adjusted indices across multiple indicators into a single score for quality reliability. This considers volume, variance and statistical modeling to render something simple, comprehensive, relevant, and actionable. There are currently millions of covered lives using this approach to direct care to specific physicians and hospitals more proficient in delivering high quality. This same approach works equally when used to assess outpatient quality. Whether having an outpatient procedure with a low number of adverse events, or an inpatient procedure where a complication is likely, this scientifically sound approach helps all stakeholders. It helps consumers move to higher quality in the best setting of care, employers paying for greater value, and providers to remediate poor quality or replicate processes which deliver high quality.
3) Avoid Unnecessary Care
Finally, avoiding poor outcomes on care that is not necessary is not high quality.Much of the current quality measurement assesses performance given the modality chosen, not determining whether the chosen modality was best. Measuring quality in the outpatient setting now brings into focus which settings, and modalities, are most efficacious in avoiding risk to the patient. Unnecessary care consumes $210 billion dollars annually. We all pay, whether premiums, co-pays, or meeting deductibles. It also exposes millions of consumers to unnecessary risk of morbidity. Appropriateness can be assessed and scored as an attribute of quality. It’s not surprising that more general surgery is performed where there are more general surgeons. Make no mistake, healthcare is currently fee-for-service. While there is much discussion around value-based care, the largest payers are still paying for all that is done regardless of whether the best method of care was chosen. Over 35 million procedures are performed in outpatient settings annually with virtually no understanding of the appropriateness, or quality performance, of the providers rendering care. This should be sobering considering most consumers have no idea of the procedural risk, or the reliability of the provider, when having surgery performed.
With the allure of much lower costs for outpatient care, proper quality assessment is vital to protect consumers, assist employers, and spur performance improvement for providers. The vacuity of quality measurement for outpatient care must be addressed for progress to be made towards value-based compensation. Quality, and value, must be discerned in a scientifically sound manner if we are to achieve the aim that Michael Porter so clearly lays out. Providers that deliver excellent, appropriate, care should benefit from their better performance as in any other market. This is currently the exception and not the rule. While there are many barriers to be addressed, the work must begin with a firm foundation. A firm foundation of comprehensive episodic risk adjustment, proper quality scoring and appropriateness of care assessment must be what the framework of achieving higher patient value sits upon to move precisely, and methodically, down the continuum. Quantros is diligently at work building the solutions where higher patient value can be measured, and rewarded, for all stakeholders.