What can a study from Zip Code 10032 teach America about Primary Care?


This JAMA study from Zip Code 10032 appears to compare nurse practitioner care with physician care. The study aimed low, intending to demonstrate no difference between physicians and nurse practitioners. As we understand more and more about social determinants, it should be surprising that any differences will be found in populations with the most social determinant limitations, such as those in this study.

This study can only teach America about a small portion of primary care for primary care delivered 1995 – 1999 in and around zip code 10032 about people that live in or around zip code 10032 some of the time. This study presented in a national journal is one of the least relevant with regard to United States primary care and United States health care.

The study uses old data, the sites of practice were different and the nurse practitioners actually changed locations during the study. Zip code 10032 is one of the most densely populated in the US with transportation and access unlike most of the US. The patient population was 80% Hispanic, neither type of provider spoke the language well, and there was little consideration of language or culture. Given the substantial social determinant considerations, it is unlikely that any variations between providers can be demonstrated as in numerous studies illustrating social determinants as major limitations of outcomes.

What will emerge from this study is controversy. This was an obvious reason for the publication. What will escape notice is whether the study was representative or even useful given the limitations of the design and the passage of 12 years since the study ended.

A major problem in health professional literature is lack of relevance. Studies can demonstrate significant differences in areas such as pharmaceutical drug treatment without being relevant. The significant differences are so small in impact such that hundreds or thousands must be treated to result in a single improvement. This often involves more costs and often does not improve overall outcomes such as better health or longer life.

Another failure in relevance is that studies commonly involve populations that are not representative of the American people. For example certain populations can be studied over and over again while others have few studies. Drug studies once again tend to involve a very narrow range of patients. Health care studies also involve a narrow range. The classic studies illustrate care of the 1 person in 1000 found in academic centers where most studies are done.

People who can access health care can be studied while those who cannot access health care and those with limited health access are less likely to be studied, particularly when using databases involving services. No services means no studies. Those studied are more commonly higher income, more urban, and more likely to have health care coverage. Those studied are often those with access or too much access. This presents a problem since America understands less to much less about populations left behind or over half of the nation.

The studies that are funded follow the health spending patterns. About 85% of the research funding goes to 3400 zip codes clustered together in 4% of the land area where the nation has top concentrations of researchers, teachers, academic institutions, information systems, health professional associations, and subspecialists. Foundation funding follows the same pattern of all lines of revenue and the top level in each line directed to zip codes inside of current concentrations of people, income, health professionals, and health spending. Physician Distribution by Concentration

What is substantially left out of awareness, understanding, and health spending is over half of the American population and basic health access areas most important to half of Americans left behind such as primary care, rural health, and care for underserved populations.

JAMA has already demonstrated that it is capable of publishing articles that comprehend the impact of social determinants of health and variations in the quality of care to underserved populations (Hong, September 2010). JAMA has also demonstrated failure in this ability as noted in a recent article about Critical Access Hospitals. This has been addressed at Which is it JAMA?

Once again JAMA presents a study in a way that appears to indicate national representation. Although the subject of the study is a comparison of primary care nurse practitioners to primary care physicians, it is important to understand what the study represents and does not represent. It represents a finding of not much difference with regard to two different types of primary care providers in a certain setting. This study also involves a location and population that is one of the most atypical in the nation.

Before you jump to conclusions about bias on my part toward nurse practitioners, I urge you to review any number of writings that have consistently avoided a quality argument. Quality arguments make little sense in a land missing in primary care where social, environmental, and political decisions have so much influence beyond a provider. Even nurses have more influence as they often encounter patients before, during, and after care and typically encounter those that are in need of better quality of care. 

As an expert in basic health access, an area largely missing for most Americans, my bias is in favor of those missing from health access. My perspective is also an objective look at what studies can and cannot demonstrate – an area I commonly find not addressed in most primary care studies involving workforce. This also comes with the full knowledge that what I note is not likely to be as relevant in ten years if we actually begin to examine health care - from more correct perspectives.

This study is also lacking in this area. JAMA once again fails to indication important study limitations in areas such as relevance for the nation, the impact of social determinants of health, and understanding the context of care provision involving the study. Zip code 10032 and surrounding zip codes are the likely sites for care as well as the locations of residence for those accessing care during the study. The population density of the area is most atypical as there are 70,000 to 100,000 people per square mile. This is up to 1000 times the population density of the nation and 100 times the 1000 people per square mile where most people reside in the United States. The study area is just inside of a bullseye of top concentrations of health workforce and health spending surrounded by populations short of health access by our national design. Zip codes east and north of this map are designated as shortage zip codes (HPSA) completing the doughnut of top concentration surrounded by shortage.

Even populations inside of concentrations lack health access as implied by the study of course very little about their health access was reported by the study.

Zip code 10032 has
  • A well developed public transportation system – important for health access and missing for most people. 
  •  A top income level (but also a top cost of living and significant numbers in poverty)
  •  A top concentration of primary care including primary care sources not listed such as training sites for MD, NP, and PA graduates. 
  •  A top concentration of health professionals 
  •  A national lowest level of primary care and family practice by percentage of the workforce (those most consistent in primary care and in breadth of primary care scope are often driven away by narrow scope and marginalization). Internal medicine and pediatric proportions are also highest in such areas - concentrations of people, income, and academics.
  • Top concentrations of subspecialists - Such areas have highest utilization rates, and paradoxically most difficult times finding specialists – specialists find a way of increasing utilization that has defeated any and all attempts at reform. One possibility is that the patients were least impacted by primary care as their care was provided by ready availability of other venues from care in other nations to self care to local support to other primary care to subspecialty to neighbors and friends. This is much less of an issue in rural locations or in settings with a complete population studied. 
  • New York has 1.39 times more physicians per person and this location is one of the highest concentrations in New York. About 85% of physicians are found in zip codes with over 75 physicians in New York. Only a few states with top health spending have such concentrations of physicians. In 2005 the one zip code of 10032 had over 900 physicians and likely has over 1000 now.
  • New York state has about 20 to 1 richest to poorest or top income quintile to bottom income quintile and the practice environment around 10032 may be more divided similar to Washington DC at 30 to 1.
  • Primary care may be most challenging for providers as relatively lower salaries also are compounded by highest cost of living.
  • Highest cost of delivering primary care (or any care) also can be a limitation on the care provided during encounters.
Some of the study limitations include


  • Over 80% of the patient recruitment involved Spanish language. There was not mention of Spanish language with regard to the patient encounters. There was a limited mention of Spanish language in the providers. The language factor was not included in the study outcome. Gender and other similar background areas have been noted to be important in measures of patient satisfaction. Patients matching up best in background to their patients tend to rate satisfaction higher and this may impact quality as well. This is a huge problem for primary care when so much of the interaction measured in an encounter is person to person.
  • Only 1316 patients is a small fraction of those who could have participated. It is a tiny fraction of the local population and of the primary care and of the overall care provided. 
  • There is not mention of what proportion of the care of the patient was delivered under study parameters. No conclusions can be made about overall care outcomes as the study was too short, too small, and too limited.
  • Health care delivered in the late 1990s during a peak emphasis on primary care under managed care influence – arguably one of the most cost effective periods in primary care in the nation’s history. Since this time less efficiency is likely (tests, referrals). The significant investment of Columbia Presbyterian was noted since 1993 in the article.
  • Nurse practitioners emphasis has changed as have the proportions remaining in primary care. Studies often imply a larger primary care role but fail to mention the limitations. Only 70% of total NP graduates are found active as direct care clinicians and only 35% are listed in primary care (HRSA Nursing Reports 2004 and 2008). Remaining in primary care is difficult for nurse practitioners as with physician assistants and physicians. Smaller portions of NP graduates train in primary care and this is even smaller compared to advanced nursing totals. Fewest years in a career also limit primary care contributions for NP compared to physician or physician assistant.
The article did not demonstrate much in the way of differences. This should not be a surprise for a number of reasons listed above and a few more. First, health care outcomes are far more likely to be driven by factors related to patients rather than type of provider. Patients or their mothers or their spouses have substantial say over encounters. For those not in the country for substantial period, there are even more outside determinants and influences. 

Does the study support quality of care as same or similar for nurse practitioners versus physicians?

Yes and no. The study statistics support same or similar care. Given the limitations of all workforce studies and failures of realistic attempts to control for important variables, it is safe to say that current workforce studies fail to have the ability to demonstrate differences in one or more key areas:
  •  Too short, not comprehensive enough, not enough isolation of the study subjects from other sources of care, lack of randomization, failure regarding controls, and failure with regard to the inclusion of many controls as indicated by Hong in JAMA and the Oregon Medicaid randomized study.
Readers can decide whether JAMA, editors, reviewers, authors, sponsors, or key stakeholders such as nurse practitioners or health care associations or health insurance foundations have an agenda. As far as this editor, author, and reviewer can state after 28 years of delivering and teaching and researching health access, I see evidence of lack of critical review by reviewers, failure of editors in editing in areas such as demanding critical review and more limitations in the published article, and widespread campaigns involving multiple articles submitted in areas such as generic expansions of physicians or nurse practitioners, the value of primary care without much justification, generic sources of primary care as solutions for primary care that are 60% - 85% not primary care, and basic health care needs of most Americans largely ignored.

And matters are getting worse not better – in directions of health spending and in the health information sources that are supposed to guide our national leaders.

Journals, editors, health professional associations, and academic institutions focus on innovative models of primary care workforce and primary care delivery but fail to focus on more primary care spending more specific to primary care services - the only real way to increase primary care delivery capacity in the nation and to improve health access and to improve the quality of primary care - mainly by moving more patients into basic health access.

Our current design with too few remaining in primary care workforce insures least experienced in primary care and also more patients moving to little or no health access and

That is a problem for cost, quality, and health access for an entire nation.

Thanks to all 12,000 who have visited Basic Health Access in 2011.

Robert C. Bowman, M.D.        Basic Health Access Web    Basic Health Access Blog

Dr. Bowman is the North American Co-Editor of Rural and Remote Health and a Professor in Family Medicine at A T Still University School of Osteopathic Medicine. He was the founding chair of the Rural Medical Educators Group of the National Rural Health Association, he was the long term chair of the STFM Group on Rural Health, he is the founding director of Priority Infrastructure at http://www.infrastructureamerica.org/ and he is the author of the World of Rural Medical Education, and Physician Workforce Studies
SMART – Specific, Measurable, Achievable, Realistic, Timely

Comments

Popular posts from this blog

Ending the Disruption of Pay for Performance

Start with CHIP to Return to Sanity

Does Academia Compromise Health Care for Most Americans?