The prescribing physician uses electronic prescribing tools that cross-check allergies and drug–drug interactions before the order hits the pharmacy queue. Radiology images are acquired and routed through a PACS viewer, not a lightbox. None of these systems act alone. They are tightly coupled-sometimes elegantly, sometimes painfully-into the clinical workflow. The Institute of Medicine’s Crossing the Quality Chasm report made this explicit two decades ago: if health care is going to be safer, more effective, and more coordinated, information technology has to be a central structural element, not an afterthought.
The most obvious point of integration is the electronic health record (EHR). In the mid-2000s, analytic work by Hillestad and colleagues laid out a bold vision: if electronic medical record (EMR) systems were broadly adopted and properly networked, the U.S. health system could save more than $81 billion annually through efficiency gains, reduced duplication, and fewer adverse events. This was not just a technical observation; it was a structural claim about how information flows: The record is no longer a static archive but a live object.
Orders, results, documentation, and billing become entries in the same information backbone. Quality measurement and research can be done by reusing the data generated in routine care, instead of running parallel data-collection infrastructures. Blumenthal and Glaser’s Information Technology Comes to Medicine in the New England Journal of Medicine took that argument straight into policy: they framed health IT as a general-purpose infrastructure for care, with benefits in safety, coordination, and efficiency-but also with serious risks if implemented poorly.
Subsequent critiques have pointed out that, in many systems, the early waves of EMR deployment produced new forms of fragmentation and burden:Clinicians now spend large portions of their day documenting for the system rather than talking to patients. Poorly designed interfaces create new error modes: picking the wrong patient from a drop-down list, selecting the wrong dose from a crowded order screen. Data silos persist when systems cannot interoperate across institutions, undermining the very continuity they were meant to support. The Crossing the Quality Chasm blueprint anticipated this risk: it argued that IT should support the six aims of quality-safety, effectiveness, patient-centeredness, timeliness, efficiency, and equity-but warned that technology deployed without redesigning care processes would not close the chasm. So the integration of IT into medicine is not a binary good. It is a powerful structural change that must be steered.
When these criteria are present, you no longer have “IT used in medicine”; you have medicine that is co-defined by its information infrastructure. That is the central observation: information technology is not sitting beside medicine; it is woven into how medicine reasons, acts, and improves itself.