Unintended consequences of the EHR

Even if the eHR is only used as a word processing and data repository - in which case it is a very expensive technology used to duplicate the paper world - the impact on work flows would be substantial. Using the eHR as a catalyst to improve care means dramatically re-engineering huge segments of the care process. Either way, unintended consequences are inevitable.

So, what might a practice or medical center do to minimize the risk of harm from unintended consequences?

It turns out that this has been discussed and studied in some depth from a fair range of perspectives. Recently, AHRQ commissioned a study by the Rand Corporation and published their conclusions as a guide to anticipating and reducing unintended consequences associated with the eHR. The report is loaded with useful information and worth careful examination.

The most common significant unintended consequences are:

  • More work. EHRs demand that providers input more (and more complex and precise) patient information. This is almost never adequately addressed. Providers uniformly report and complain about documentation time taking away from clinical time.
  • Dysfunctional workflows. IT design may streamline specific processes but often at the cost of disrupting clinical flow and with the disruption or loss of other clinical systems. There is also a tendency by non-clinicians designing and implementing eHR systems to expect consistency and conformity because they are unaware of the extent and value of diversity in clinical medicine. This has been humorously described mathematically: (crap) + (technology) = fast expensive crap
  • Perennial demand for upgrades. EHRs are all in constant beta. We are designing tools for a world that does not yet exist. The knowledge upon which medicine and medical practice is based is changing at an incredible rate, so best practices and ideal workflows are in constant flux. New technology and business changes mean that new models, designs, interfaces and functionality arrive every month. This is a huge undertaking, and the personnel and budgets managing this generally live in the IT and financial world and are rarely clinicians.
  • Change is hard. Most experienced clinicians have developed and are comfortable with cognitive and workflow approaches based in the old but tried-and-true paper world. Changing deeply ingrained thinking processes and behaviors is difficult even when it is clear that the change will be beneficial. Because not everyone changes at the same speed, and some things canot be changed, one often has multiple simultaneous and not easily compatible systems and world views in place.
  • Different communication patterns. The paper-based model evolved in a slower world dependent on direct, verbal communication. A transition to EHRs requires simultaneously development of a new culture and new forms of communication and collaboration, not just within the eHR, but broadly within the institution. This is rarely recognized or addressed.
  • Poor user adoption. The practice of clinical medicine is a demanding task with very high stakes. Users need both an in depth understanding of the new systems and a solid belief that the painful transition will result in better and easier medicine. This cannot happen in the usual setting where the clinicians are not involved in setting criteria and goals, establishing priorities, selecting the system(s), modifying the system to meet specific clinical needs, designing the process of roll out, and monitoring for success, failure and the need for change. Clinicians do this for a living and are rarely comfortable being passive voyeurs and passengers in a process that has such a huge impact on their work and the well-being of their patients.
  • Categorically new errors. The greater the change in systems and workflows, the more likely one is to create novel errors, which are - by definition - impossible to anticipate. Worse, there is rarely a recognition of this danger or a system in place to address these issues when (not if) they occur.
  • Changes in role, and loss of autonomy. The more powerful the eHR and the larger the investment, the more disruptive the change from the older patient-clinician model where the clinician functions with considerable autonomy to the newer patient-system model, where the clinician spends a great deal of time functioning as an interface between the patient and the database rather than as an autonomous collaborator with the patient. Clinical departments and clinical cultures lose power in the organizations as they are required to adhere to eHR guidelines designed and enforced by IT and quality departments and are expected to maximize productivity to amortize the investment in the eHR and the growing cost of the IT and quality departments.
  • Atrophy of clinical skill and replacement of clinical judgement with algorithms and protocols. Built-in electronic point-of-service clinical support has the potential to create lapses in a doctor's knowledge base, a lazy acceptance of the algorithm, and a cultural change away from patient-centered care to rule-driven care. This is augmented by institutional emphasis on evaluation by metrics and targets, and quality defined as standardization and homogeneity.


I have seen varying degrees of all these problems in my own institution over the last decade. How about you? And what are you doing about it.



Links to more on this topic::