Generalizations and Assumptions Dominate Innovative Health Care Design: The Flawed Literature Basis

Robert Bowman Robert Bowman

Robert Bowman

Basic Health Access

发布日期: 2020年9月14日
+ 关注

Would it surprise you that those who design health care have failed to improve it time after time for decades despite numerous reforms? Costs rise, outcomes are declining, and health access is worsening. The designers have focused upon micromanagement of costs and of outcomes - which make the situations worse. This should not be a surprise given the flawed literature that continues to support meaningless, costly, and burdensome interventions.

Few consider that health care designs are based on generalizations and assumptions across policy and research.

The designers are not held to "DO NO HARM". They are given free reign to experiment upon human subjects - the American People. They can abuse vulnerable populations and they do just exactly that with their designs.

Designers join a chorus of voices that cry out for an evidence basis for health care delivery while not following scientific evidence basis in their various innovative health care designs that hope to disrupt health care and replace it with a better design.

  • Disruptive is a particularly apt description since what they are constantly doing is undermining the team members delivering the care - particularly where most Americans have lowest concentrations of health care workforce.

The Contrived Medical Literature

There is a general designer support for innovation, regulation, and digitization. This bandwagon was set in motion in the 1980s, fought conflicts in the 1990s, and has rolled over much of the nation since that time. Few consider that they are worsening costs, quality, and outcomes by their designs.

Once you begin to examine the redirection of health care dollars away from most Americans and toward designers and higher concentrations - you can begin to see disparities and outcomes worsened by design. Obamacare is a great example. The design appears to focus upon quality improvement and better access thru insurance expansion, but the billions stolen from the practices and populations most behind testifies to a massive transfer of dollars from lower to higher concentration populations. And only about 10 cents on the insurance dollar returns to the local practices and hospitals already most behind.

The journals have generally considered micro-management to be progressive and innovative. And sadly the publication choices are often about studies with positive findings. This is a particularly bad choice for the type of research done.

  • Government and foundation sponsors want to demonstrate progress and fund projects that are supposed to support interventions.
  • Researchers depend upon grants that are focused upon the interventions and know that they grant must demonstrate positive results. The way that the design, collect, and process data has much to do with what is discovered and reported.
  • Government programs have often had vague definitions and variability in the variables measured. And the metrics and measurements can change over time.
  • And even when the research seems to contradict progress with innovation (as in Primary Care Medical Home) - this critique is minimized.
  • Sometimes the authors continue to study their work after publication - and find that the way that they processed the data contributed to the findings. These authors were brave enough to follow up and do a retraction when they found out that there intervention actually had consequences rather than the benefits initially claimed. "The initial study by Aboumatar et al3 was a randomized clinical trial of a 3-month intensive chronic obstructive pulmonary disease (COPD) intervention that combined transitional care support and chronic disease self-management. https://jamanetwork.com/journals/jama/fullarticle/2752449

Low Level Scientific Validity

Note that these research excursions do not have the highest level of scientific validity. They are generally not randomized controlled trials. They use convenience databases that include much about billing and little about the patient other than their comorbidities.

The research often involves regression equations that can demonstrate association but not causation. But sadly these research explorations are often allowed to look causative due to the lack of limitations and the failure to explain away alternative hypotheses - such as the differences reported in the study were really about differences in the populations being compared.

The journals fail to require the authors of their studies to reject the null or alternative hypotheses.

The media and health care journalists fail to critique this well and add their own distractions.

In general the populations involved in clinical interventions involving costs or quality are populations with better social determinants and other factors - a likely reason for the positive findings. Sometimes this is easy to see as with comparisons of

  • Rural to urban hospitals
  • Low to high volume hospitals
  • Primary Care Medical Home compared to IPA
  • Studies involving Kaiser or Geisinger or Washington State

Or you can see flawed research in studies comparing NP to MD where the similar outcomes are likely do to similar populations compared and not NP vs MD. This study from Michigan was particularly bad with poor controls, old data collected for other purposes, a change of clinic site for NP, and more.

A great review of numerous policy areas can be found in this compilation of the works of Kip Sullivan. https://thehealthcareblog.com/blog/tag/kip-sullivan/

Regressive Regression Use and Abuse

These excursions that are suitable only for exploration are predominantly based on regressions that assume that means are accurate. As Jha and others point out, research methods that use means for data points only represent about 10 - 20% of the data closest to the mean depending upon the Bell Curve pattern. Therefore each data point has substantial inaccuracies. And we know that humans are more complex and much of the data is missing regarding human interactions, behaviors, health literacy, and much more.

  • "What is the central tendency of a distribution but a lazy generalization? The aggregate, the mean, is wrong about everyone but the few closest to the mean, yet is so revered because we mistake the aggregate for the truth. The tyranny of the aggregate is the most extraordinary tyranny of our times. The aggregate is built by people who vary, yet it imposes itself on the individuals, the very variation which creates it. It literally bites the hands that feed it."  SAURABH JHA, MD (associate editor with The Health Care Blog)

In other words, to do good research you must actually begin and end with substantial original data collection and processing specific to the subjects and the intervention.

  • How did we ever tolerate research that is about number crunching and not about people understanding?

Convenience data and retrospective studies can be processed and reprocessed in ways that make the interventions look good - but have too many flaws.

Researchers should do mixed method research or qualitative research to gain awareness with regard to the population and intervention.

The Error in Medical Error Studies

Would you believe that the medical literature would publish a dramatic study to gain attention while ignoring the fact that it was based on assumptions that relied on other assumptions previously published? BMJ defended its medical error publication by saying that it needed to be discussed. But what about science? What about the large bulk of people, patient, family, and other factors other than health care factors?

Surely performance based incentives work to improve outcomes and diminish costs - wrong.

Race, Ethnicity, and Comorbidity Shape Poor Outcomes - Maybe, But These Are Highly Associated with Social Determinants and other Variables

Race and how people are impacted by race clearly varies. Comorbidities are known to be seen at higher levels for the Americans most behind in multiple social determinant and other factors. Medical literature has a medical focus, and not a psychosocial focus. The biologic dimension impacts are small compared to the complexity of the human experience - and numerous degrees and dimensions of differences.

Studies that are particularly bad use proxy variables based on patient or person zip code. These studies may have the ambiguous variables for race/ethnicity in addition to comorbidities. Those familiar with regressions understand that these variables are pushed into significance by the lack of including social determinant and other variables specific to the patient.

As a result, bad regressions are made worse. So we focus on race, ethnicity, and comorbidity which we can do little about - while social determinant and other personal and societal factors are not included, are too few, and have not been studied well.

And we delay recognizing that we must invest in people to actually change outcomes - rather than technology, clinical interventions, clinical process, reorganizing patient medical homes

Clearly the major reason for health, education, economic, and societal failure in the US is about the failure to invest in most Americans most behind. As our designs concentrate more and more in the hands of fewer, most suffer.

In medicine it is not ethical to divert a patient to a treatment that is ineffective or may cause harm. The literature is diverting our nation to treatments that are costly, burdensome, and ineffective.

  • Meaningless Use and Abuse Sums Up Policy Designs

In the Ivory Tower world of the designers, they can continue to believe that they are helping. But they are not helping most Americans who are most behind.

Examples of Studies with Researchers Who Are Aware

  • "The 3 illustrative cases reveal it is possible for some primary care practices to seize ownership of their care and prioritize their craft of family medicine. These practices began with their founders' realization that matching their practice to their values was impossible, given the conventional financing system and commercial EHRs designed to serve it. They came to this conclusion differently but took similar action by developing business models that circumvented the limitations of fee for documentation and pay for performance. Although their clinical care and business models differed, all 3 practices succeeded in shifting the source and directional emphasis of change from outside-in to inside-out. Shifting Implementation Science Theory to Empower Primary Care Practices William L. Miller, Ellen B. Rubinstein, Jenna Howard and Benjamin F. Crabtree The Annals of Family Medicine May 2019, 17 (3) 250-256; DOI: https://doi.org/10.1370/afm.2353 https://www.annfammed.org/content/17/3/250.full.pdf+html
  • Small and medium size practices are more likely to be disrupted by changes in key personnel, EHR, billing, location, ownership, and other changes. These can be costly and can contribute to inability to adapt to any number of changes. The Alarming Rate of Major Disruptive Events in Primary Care Practices in Oklahoma James W. Mold, Margaret Walsh, Ann F. Chou and Juell B. Homco The Annals of Family Medicine April 2018, 16 (Suppl 1) S52-S57; DOI: https://doi.org/10.1370/afm.2201 https://www.annfammed.org/content/16/Suppl_1/S52

If you do not understand most Americans, you should not design health care or education for them.

要查看或添加评论，请登录

更多精彩文章

社区洞察

其他会员也浏览了