Haven’t been posting here for a while as I’ve been too busy with other more urgent issues. But I’m back, with a new category in which I have quite a lot to say.
I’ve been consulting in Business Intelligence systems for over twenty five years, and it always seemed to me that the health care industry is greatly underserved in this field. A BI system requires a store of electronically accessible data that can be manipulated to produce measures of performance for both controllable activities and non-controllable results. The value of BI is in demonstrating correlations between the two, from which decisions can be made about how best to allocate enterprise resources to achieve the best business results. By that criterion, health care is a perfect industry in which to apply BI techniques, except for the fact that most health care data is not readily accessible in computer databases.
This year, the American Renewal and Reinvestment Act of 2009 (the stimulus bill) offered incentives to health care enterprises that do business with Medicare for “meaningful use of certified electronic health records”. Depending on the nature of a medical practice, the plan could pay up to $40,000 per practitioner over five years beginning in 2011, enough (it would seem) to offset a significant portion of the investment in an EHR. The estimated time required for a practice, a clinic, or a hospital to convert from paper charts to EHR is about nine to eighteen months, so the stimulus plan aims to get practitioners to start the process now.
So far, it is a losing proposition. Physicians have resisted adopting EHR for years, for several very good reasons.
First and foremost is the cost. It’s a healthy chunk of change, despite the range of available choices (including ASP/SAAS offerings) for the initial investment, and more is likely to be required in the future under normal systems life cycle conditions. And while the incentives in the stimulus plan sound generous, in reality few practitioners will collect anything close to the authorized limits. The plan allows doctors to bill Medicare an additional 15% on each transaction, but Medicare reimbursement rates for routine transactions are so low this premium never adds up to much.
Second, data privacy issues weigh heavily. While it is true that paper charts are awkward to access and work with, they are also for the same reason intrinsically more secure than electronic records. Data security is a significant HIPAA problem that most health practitioners are not very comfortable with.
Third, practitioners are skeptical about the ability to share data electronically with other practitioners and third party payers who use systems different from their own.
Finally, most practitioners are not very comfortable with using a computer in normal clinical operations. They keep notes on paper without thinking, but if they try to do the same thing with a computer they feel that it interferes with patient communication.
I will address the first three of these objections in future posts. The issue of integrating a computer into a medical encounter is a workflow problem that I address with clients in an implementation project. I can say there is no simple solution, and what works for one person might not work for another, even within the same practice. But this is not to say it is intractable. Practitioners should not dwell on it as a reason not to adopt EHR.
Two other hurdles exist that I have not seen addressed in any discussion. One is the problem of displacement, and the other is comparative effectiveness.
“Displacement” is a term for job loss due to technological innovation. EHR automates tasks that currently require two or three full time employees in many medical practices: medical transcription and E&M Coding.
Most doctors have illegible handwriting, largely because of the conditions under which they have to write. They make notes on the patient’s chart in the examining room, and they have other patients waiting. They jot down what they have to as quickly as possible, and that information has to be transcribed so the next person who has to use it can read it. The E&M Coder uses that information to apply the evaluation and maintenance codes to the chart so that the encounter can be billed to the patient and/or the insurer.
EHR substantially reduces the need for both transacription and manual coding. The doctor enters his notes directly onto the electronic patient record either with the keyboard or using handwriting or speech recognition tools in the system. He can later review and update the notes; no other transcription is required. Most systems also automatically apply the E&M codes based on the “discreet data elements” on the chart, and these too can be reviewed and updated by the practitioner based on his notes. So a full time coder is redundant.
If you run a small business, and you know all your employees, and you have to let some go because their functions are no longer necessary, it’s very hard. Frankly, I don’t think most practitioners want to deal with that. It makes EHR a very tough sell, despite its significant benefits.
“Comparative Effectiveness” is compares treatments of specified conditions and their results across many cases. “Pay-for-performance” incentives from insurers and Medicare essentially use comparative effectiveness studies to impose treatment and reimbursement guidelines. EHR provides support for this by making it much easier to collect and aggregate data, and while “pay-for-performance” suggests significant incremental revenue for medical practitioners and institutions, it also means that practitioners can be graded like any other commodity in a competitive market.
The idea of comparison shopping for a doctor the way you do for a car would have been unthinkable when I was growing up. Nowadays, we have all kinds of consumer reporting, such as the annual best doctors report in New York Magazine. But these comparisons are subjective, based more on patients’ relations with their doctors than on any objective measurement. Because EHR can produce objective comparisons, insurers may use them to rank doctors performance. Rankings can be displayed in the insurers’ directories (both on and off line) and can impact the practitioner’s business accordingly.
Frankly, I think that is scaring the hell out of the medical profession, not to mention drug and medical device makers (see the story in the New York Times of May 7, 2009: “New Effort Reopens a Medical Minefield” by Barry Meier ). Comparative effectiveness data can be used by patients, insurers, and governments alike to second-guess the doctors’ decisions, and that is not good for the doctor, the patient, or health care in general.
The thing is, adoption of EHR is going to become mandatory, very likely before 2015. Other third party payers besides Medicare are going to want it to reduce their processing costs. And the kind of decision support information EHR makes possible is going to be critical for health care reform. The cost of health care in this country is out of control, mainly because of the way the health insurance industry does business. Collecting objective information about medical practice and patient progress will ultimately force the insurers to adopt more reasonable positions on reimbursement policy. Health practitioners need to start viewing EHR as a tool for combating bureaucracy, as well as a tool for clinical management.