By MARK BRAUNSTEIN, MD
Back in the ‘stone ages’ when I (an MIT grad) was an intern, I was called at 4 AM to see someone else’s gravely ill patient because her IV had infiltrated. I started a new one and drew some blood work to check on her status. When the results came back (on paper) I (manually) calculated her anion gap. This is simple arithmetic but I had been up all night and didn’t do it right.
She died.
On morning rounds the attending assured me that there was nothing I could have done anyway but, of course, in other circumstances it could have made a difference and an EHR could have easily done this calculation and brought the problematic result to my attention. My passion for EHRs and FHIR apps to improve them really traces back to this patient episode I will never forget.
My criticism of the recent Kaiser Health News and Fortune article Death by 1000 Clicks is generally not about what it says but what it doesn’t say and its tone.
The article emphasizes the undeniable fact that EHRs cause new sources of medical error that can damage patients. It devotes a lot of ink to documenting some of these in dramatic terms. Yes, with hundreds of vendors out there, the quality of EHR software is highly variable. Among the major weaknesses of some EHRs are awkward user interfaces that can lead to errors. In fact, one of the highlights of my health informatics course is a demonstration of this by a physician whose patient died at least in part as a result of a poor EHR presentation of lab test results.
However, the article fails to pay equal attention to the ways EHRs can, if properly used, help prevent errors. It briefly mentions that around a 60% majority of physicians using EHRs feel that they improve quality. The reasons quality is improved deserved more attention. The article also fails to discuss some of the new, exciting technologies to improve EHR usability through innovative third party apps and he real progress being made in data sharing including patient access to their digital records.
The article acknowledges that “medical errors happened en masse in the age of paper medicine, when hospital staffers misinterpreted a physician’s scrawl or read the wrong chart to deadly consequence, for instance.” It misses the opportunity to quantify the awful scope of this problem as the Institute of Medicine (IOM) did in 1999 when it found, based on 1984 data from physician reviews of New York hospital patient medical records, that “as many as 98,000 people die in any given year from medical errors that occur in hospitals.” A 2013 Journal of Patient Safety article reviewing four studies done in 2008-11 indicated that things were actually worse: “a lower limit of 210,000 deaths per year was associated with preventable harm in hospitals”. There is controversy about the exact numbers but little doubt that they are significant. The two studies cited were, of course, well before most of the wide adoption of EHRs funded by the federal HITECH program to fund EHR adoption. Moreover, while the issue is far from settled, there are studies that suggest that hospitals “with automated notes and records, order entry, and clinical decision support had fewer complications, lower mortality rates, and lower costs.” Other studies suggest that EHRs improve physician-patient communications.
Atul Gawande’s excellent article in the November 12, 2018 New Yorker describes the issues physicians have with EHRs in painful detail but it also points out that at his hospital “While some sixty thousand staff members use the system, almost ten times as many patients log into it to look up their lab results, remind themselves of the medications they are supposed to take, read the office notes that their doctor wrote in order to better understand what they’ve been told.” He lists specific quality initiatives his hospital has launched now that its records are digital. These include identifying people who have been on opioids for more than three months and patients who have been diagnosed with high-risk diseases like cancer but haven’t received prompt treatment. He also notes that the ability to adjust protocols electronically allows changes to occur far faster as new clinical evidence comes in.
In addition, the KHN/Fortune article fails to note that a substantial part of the tremendous increase in physician documentation is not caused by EHRs but by the need to obtain more data on the reasons why expensive tests and procedures are being ordered in an attempt to reign in out of control health care costs and for quality measurement and reporting purposes to support new ‘value-based’ reimbursement of care. Clearly, EHRs could be better designed to help with these burdensome new requirements and there are innovations that attempt to do this, but the article gives the impression that EHRs are the sole cause of physician burnout due to added documentation. This is simply not true.
Finally, the article gives no hint at the enormous progress that is being made to fix interoperability so that EHRs can share data meaningfully. It fails to mention that new and innovative third party developed capabilities can now be added to many major EHRs and that, because their records are now digital, many patients can seamlessly access them in support of their own health.
A more balanced article would have pointed to the 21st Century Cures Act and its mandate that EHRs provide patient-facing APIs. APIs are a standard way of retrieving data from a computer. You use them routinely to access data via the Internet and your phone apps use them to access your location or contacts. The patient health APIs will be required to use the new Health Level 7 FHIR® standard according to a proposed rule recently posted for comment by the Office of the National Coordinator for Health Information Technology within the Department of Health and Human Services.
Importantly, FHIR is the basis for SMART®, an EHR agnostic app platform developed with federal support at Harvard’s Boston Children’s Hospital. Using a SMART app, I was recently able to download my digital health record from my health system in less than 5 minutes completely on my own. The only specialized knowledge required was that I knew it could be done. After registering I simply selected my health system from a list and input the same portal credentials that once gave me view-only access to my own health data. Having done that, I now could use other SMART FHIR apps of my choice to manipulate my data in whatever way I find useful.
Using this same API, Medicare now offers its 58,000,000 beneficiaries access to their claims-based health data using apps of their choice and the Veterans Administration now offers its patients similar capabilities using their EHR data. Of perhaps even greater significance, Apple has embraced this new API so iPhone users can aggregate their EHR data even across providers. I know two people who have recently done just that on their own using their portal credentials. Once aggregated the EHR data is available to innovative iPhone health apps along with any data the patient may have collected on their own. Apple is actively supporting developers in creating those new health apps.
Progress is not just about patients. Unmentioned in the article are the new SMART FHIR app galleries operated by several of the major EHR vendors discussed in the article. Many apps in those galleries are written by innovative new companies and are generally for physicians and other care providers. Some of them directly address usability issues discussed in the KHN/Fortune article.
CMS administrator Seema Verma is not just talking about fixing these problems, as the article implies, she has called for the use of FHIR for patient, provider and payer applications. The payer community has created the Da Vinci Project, an ambitious collaboration to define an initial set of use cases some of which aim to streamline and largely automate their communication with physicians for preapproval of tests and procedures and quality reporting.
Yes, the health care system has huge issues with patient safety, but it had them before EHRs and there is no evidence I am aware of that they have, in aggregate, made the problem worse and there is evidence that they have made new quality improvement efforts possible. Moreover, unlike paper records, digital health care records can and are being shared. The data they contain is becoming the platform for innovation that is so badly needed to fix EHR usability and efficiency issues and many of the other structural problems in our massively complex, error prone and expensive US healthcare system.
The article got a great deal of attention, so it is regrettable that it is a missed opportunity to give a more nuanced and balanced view of digital healthcare today.
Mark Braunstein, MD developed one of the first electronic medical record systems in the early 1970s. His latest book is Health Informatics on FHIR: How HL7’s New API is Transforming Healthcare.
