[Draft]

"We must find a way for small practices and rural and underserved providers to participate in exchange." says Brian Ahier of the Oregon HIE Planning Team.

While NHIN Direct was formed to provide "an easy 'on-ramp' for a wide set of providers and organizations" according to the home page, there is particular interest in seeing it support the needs of such providers. This page is about building a collective perspective on what those needs are. This perspective should be useful not just in discussing what protocols make sense on the backbone and at the edge, but also what the reference implementations will need to do to support the needs of the pilots. We also need to spend a little bit of time talking about the business realities at the small provider level.

What is a "Small Provider Office"? Defined in this document of a practice setting of 5 or fewer providers. Approximately 42% of all office physicians work in practice locations of this type, skewed towards 1 and 2 provider practices. It is useful to validate these assertions with a reference-based "Environmental Scan" of the NHIN Direct Ecosystem and calculate the impact on patients served by these physicians.

NHIN Direct Health Ecosystem Environmental Scan:

Executive Summary:

Target Population Summary:

The NHIN Direct Project specifically targets extending health communication specifications that support the subset of the provider community (and thereby support their associated patients) that:
  1. Are NOT employed by organizations, i.e. the individual health care providers specifically targeted by NHIN Direct are full- or part-owner of their provider practices
AND
  1. Do NOT subscribe to EHR services provided by an ASP via the ASP data center
AND
  1. Serve rural patients or under-served urban patients

Associated Assumption Summary:

  1. Organizations owned by non-physicians that also are large enough to employ physicians are large enough to purchase IT services for their own provider practices (and therefore have managers rather than physician-owners making the final IT decisions). These organizations may also install one of the inexpensive NHIN Exchange Appliances on the market (whether they utilize NHIN CA certs or or other CA certs) for organization-to-organization NHIN Exchange-type provider communications. These larger organizations still need to communicate with small organizations (usually defined as 1 to five physicians) who refer patients to them.
  2. ASP services imply limited technology requirements at the provider site that eliminate the need for NHIN Direct specified-technology at the local provider site.
  3. Well-served urban patients have plenty of choices on whether to utilize electronic or non-electronic provider practices for their care, and market forces will determine physician IT choices in urban areas, i.e. NHIN Direct specifications are simply an IT option for these well-served patients and their providers who don't fall into an “IT hardship category.”

Environmental Scan Summary:

Of the approximately 750,000 physicians in the US, 450,000 physicians or less provide patient care in physician offices (see “Total Number of US Physicians” below). Of these, about 190,000 physicians (42% of 450,000) provide care in practices of one to five physicians (see Distribution of Physician Practice Size). Since the percentage of primary care owners in practices is higher than specialist owners of practices (see Physician Ownership), referrals of patients from these small primary care practices (<6) to other small specialty practices (assuming a random distribution of referrals) can be no higher than about 20% of patients in the US (42% X 42%) as an urban-rural average. By subtracting from this 20% of the US population, the number of patients in these small practices supported by ASPs, and by subtracting the number of well-served urban patients who have many choices of electronic or non-electronic practices, one can estimate the number of patients who will benefit from NHIN Direct specifications. Therefore, one can assert that some percentage of patients very much smaller than 20% of the US rural and urban population from these 100,000 primary care physicians depend on full source-to-destination, small practice-to-small practice, NHIN Direct referrals. Of course, the very large majority number of patients served by the small practice primary care physicians will depend on NHIN Direct small practice-to-large practice referral communications. Finally, the remaining population of patients from the slight majority of primary-care physicians who are employed in practices with greater than five physicians (50+%), depend on large practice-to-large practice referrals, which may not require NHIN Direct specifications at all.

Environmental Scan Qualifications:

  1. Statistics drawn from different sources vary in actual numbers. Consequently, utilizing fewer significant digits in the concluding data rather than individual data sources is prudent.

Environmental Scan Source Data:

Total Number of US Physicians:

The Department of Labor ( http://www.bls.gov/oco/ocos074.htm ) identified 661,000 US physicians and surgeons (2008), of which 77,000 (12%) were self-employed and enjoyed a higher income than employed physicians. Of the employed, 349,000 were employed by “Physician Offices.” As is noted below in Group Practice Distributions, a portion of these employed by “Physician Offices” are also part-owners of the “Physician Office” group practice.

The AMA count of physicians (~800,000) is thought to be somewhat high due to physicians who are actually retired still counted on their lists:
JAMA. 2009 October 21; 302(15): 1674–1680. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2791886/

Number of US Physician Conclusions:
The correct number of total physicians in the US is probably between 700,000 to 750,000 with at most 60% of these physicians (349,000 + 77,000 = 456,000) either self-employed or employed by “Physician Offices,” in which they might be part-owner.

Distribution of Physician Practice Size:

The trend overall is for physicians to move from 1-5 physician practices into larger, single specialty group practices (6-50 physicians):
( http://www.hschange.com/CONTENT/941/#ib1 ). Table 1 from this source:
Physicians_By_Practice_Setting.png

At the same time that ownership (full or part-ownership) in primary care practices are being retained by primary care physicians, specialists who receive referrals are relinquishing ownership in practices:
Ownership_by_Physicians.png

Physician Distribution Conclusions:
About 42% of physicians practice in office-based practices one to six physicians in size (32% + 9.8%). The trend is for about 50% of physicians in primary care practices to continue to own all or part of the primary care solo and group practice, which means that they will continue to make their own IT investment decisions, including EHR decisions. This means that only about 50% of all patients are affected by the IT investment decisions of their primary care owner-physicians including those who practice in large group practices. At the same time, the trend is for the majority of physicians receiving referrals, most of whom do NOT own their own practices (<50%), also do NOT make their own IT decisions. Therefore, a crude estimate of patients referred from a primary care owner-physician to a specialty care owner-physician might roughly be 20-30% of patients in the US (50% X 50%), assuming random referral patterns. This figures are extremely important to judging the frequency and impact of source-to-destination, end-to-end, referral communications by NHIN Direct. Similar logic using the statistics reported above would conclude that practices of less than six primary care physicians referring to practices of less than six specialty physicians might be roughly 15% (42% X 42%) of patients in the US, whether rural or urban.

Rural and Urban Distributions of Primary Care and Specialty Physicians:


Despite the smaller number proportionally of under-served patients in the US (rural and and under-served urban populations) compared to national population figures, a larger-than-expected proportion of the poor US health statistics are traced to these under-served populations. Consequently, there is major concern to promote IT strategies that support providers (including physicians) who serve both rural patients and under-served urban patients (those with various disparities from the average urban population). Therefore, it is useful to examine the distribution of physicians by specialty between rural and urban settings: http://content.healthaffairs.org/cgi/content/full/22/1/190/F5

The trend is for counties with less than 100,000 population to contain 2-3 generalist/FP physicians per primary care internist (total about 40 primary care physicians per 100,000 population). As the county population rises above 200,000, the GP/family practice numbers per 100,000 population remain stable, but a higher proportion of patients receive care from primary care internists, generalist obstetricians, and generalist pediatricians (total about 100 primary care physicians per 100,000 population). As an average state, California mimics these national figures (West J Med 1998; 168:412-421). At the same time, in California, the specialist physician statistic was 120+ specialist physicians per 100,000 population in urban areas.

Rural Distribution Conclusions:
Referrals from very rural areas to specialists occur largely from generalist/family practitioners to urban specialists with about a third from general internists to urban specialists. Urban referrals occur from a more even mix of primary care physician-types to urban specialists as well as from urban specialist to urban specialist. As noted above in “Distribution of Physician Practice Size,” the tendency is for medical specialists in general to cluster into group practices with less practice ownership by specialists (53%). This would indicate that urban-to-urban referrals for the majority of patients are dominated by groups of primary-care and specialty physicians to other groups of more than six specialty physicians, most of whom are not owners of the practice and do not make ownership decisions on IT investments. The conclusion is that that urban referral communications between small groups and small groups is far in the minority, whether under-served populations or not, and certainly represent no more than 15% of the urban patient referral population. At the same time, one could postulate that rural-to-urban referral communications from small-group to small-group could represent larger than 15% of the referrals for rural patients. This conclusion would need to be tested with real surveys. This statistic could easily be swayed by marketing efforts of larger group practices. In any case, although the numbers of referrals from under-served populations, whether rural or urban, to small, urban, specialty practices need additional study, the proportion is not large, probably below 25% of all referrals for under-served populations.

Small Provider (<6 providers) Technology Environment


  1. Electronic billing system, either on-premise or ASP
  2. Consumer-grade operating system and hardware - likely majority MS Windows (XP and Vista), very small numbers of Mac clients. Predominantly used from desktops, more often for use by administrative staff than clinical staff
    1. Providers often have smartphones
    2. Some use of laptops or use of tablet-style and mobile devices, typically at practices that already have an EHR or eRx on-premise or ASP.
  3. In some cases (20-40%), some degree of charting technology, split half between use of "complete" systems and "modular" systems (the latter consisting often of standalone eRx, or lightweight charting systems or ASP).
  4. Some informal information sharing with other doctors and patients using ordinary e-mail.
  5. Of the situations involving ordinary e-mail, some use an ISP-based email and some use a hosted email service like Yahoo, Hotmail, or Gmail. Rarely would any practice with less the six physicians host an email server.
  6. Small providers face significant IT management hurdles in adopting on-premise software, either client or server software, for managing health data;
  7. Even without considering technology hurdles, small providers face the same operational and workflow hurdles in using a new technology to create, store and manage patient data of which large providers complain.

Conclusions:

  1. NHIN Direct should avoid mandating new on-premise software as much as possible.
  2. NHIN Direct should avoid mandating solutions that cannot be integrated into other workflows.
  3. NHIN Direct should, however, encourage and be a path forward for use of Certified EHR Technology.
  4. Small providers would face a modest hurdle in using new web-based software for messaging other doctors and patients. An ideal solution would allow doctors to continue to use the same messaging tools (Outlook, Thunderbird, etc) and the same messaging websites (Hotmail, Gmail, their ISP) they use today, but it would be acceptable to ask them to use a new web-based messaging environment.
  5. Small providers would face a few hurdles configuring their existing messaging tools to use new servers - that is, a new POP/IMAP server from within Outlook. Providers should be expected to be able to use the correct account to send PHI-confidential messages to other providers or to the patient.

Other Business Realities

  1. Small providers are already overstretched by the technology they are asked to use. NHIN Direct must present a low barrier to use in not just deployment and maintenance costs, but also in training, fixed costs, end-user complexity, and clinic-level complexity.
  2. At most, some very small percentage of these small providers are comfortable running their own messaging servers - be it SMTP servers or otherwise - so long as such tools are commercially supported and robust enough to impose little maintenance costs. Those small providers would rather run these servers than pay a monthly fee for a hosted secure messaging tool. In NHIN Direct parlance, these small providers might want to be their own HISP. A survey should determine this percentage.
  3. Some percentage of small providers might be comfortable paying a monthly fee to a hosted HISP service. Today, most use free messaging services from ISP, Google, or Yahoo. Those small providers might pay a small amount ($?) to move to secure health messaging and achieve Stage 1 Meaningful Use. A survey should determine this percentage.
  4. Technology providers to small providers operate in a lean environment. They sell licensed products for low per-clinician fees (generally $5,000-$30,000 for EHR-level software, or $$?? for simple messaging or e-prescription services) or work on a consulting basis. The consulting rates are at a $$ level that requires the technology be relatively simple to deploy and train against. {I don't think consulting based solution are targeted to small providers, except open source which is not widely deployed}
  5. Many of the providers will not have access to highly structured, codified data with extensive meta-data. Providers need to be able to exchange a number of different common formats of data between them. Some will be able to exchange structured data, some will be able to only exchange text or PDF.
  6. Many of the providers that have been early adopters of EHRs will need to make another investment to procure the capabilities to achieve meaningful use.