Population Health Management Case Study: Kaiser Permanente

Electronic Clinical Surveillance for Identifying Care Gaps

Magnifying glass
Creative Crop/Digital Vision/Getty Images.

Electronic health records (EHRs), patient registries, and other health information systems are valuable tools for population health management. The basic concept is to scan electronically available health data for warning signs to identify patients who might benefit from extra attention from the health care team.

This case study will highlight a real-life example of how a health care system, Kaiser Permanente Southern California (KPSC), conducts electronic surveillance of health information to identify opportunities to improve the quality of health care (“care gaps”) in the outpatient setting.

The salient features of KPSC Outpatient Safety Net Program were described by Kim Danforth and colleagues in a 2014 article in the eGEMs (Generating Evidence & Methods to improve patient outcomes) journal.

Guiding Principles

A few guiding principles of the KPSC Outpatient Safety Net Program are worth noting. First, the program is based on the premise that care gaps exist despite best efforts to provide high-quality health care. If 990 out of 1,000 patients with abnormal lab results receive appropriate follow-up, that still leaves 10 patients who would fall through cracks.

Second, the overall approach is to conduct electronic surveillance of clinical data apart from the actual provider-patient encounter. Just as the name implies, the program functions as a safety net to “catch” patients whose problems might slip by unnoticed during the course of a busy clinic session. KPSC used their Epic-based EHR to monitor lab results and medication prescriptions for red flags.

Third, the purpose of the program is to identify patients who need extra attention or follow up care, rather than evaluate performance of individual physicians. The “blame-free, safety oriented” nature of the program likely increased acceptance by health care providers.


Leadership used the following criteria to determine if an individual safety net program would be developed.

  • Clinical Impact: A safety net program was not deemed worthwhile if it did not address a significant patient safety issue.

  • Identification: The care gap needed to be easily identified through structured data in the EHR, such as lab values or diagnosis codes. Filtering structured data is much easier than conducting natural language processing of free text notes.

  • Follow up: Adequate personnel and resources were required to follow up with patients. If a large number of patients would be affected, automated solutions were sought (e.g. automatically send reminder letters to patients)

Individual Safety Net Programs

A total of 24 safety net programs were described in the article, with each focused on a specific care gap. Programs were established to aid the diagnosis of cancer (cervical, colorectal, prostate), chronic kidney disease, and hepatitis C by improving the detection and timely follow-up of abnormal screening tests.

Other programs monitored for potential adverse effects of medications. This was achieved by identifying abnormal lab values that would suggest that drug levels were too high or were harming the kidneys, liver, or other organs.

The medication monitoring programs also identified patients who were prescribed potentially harmful doses of individual medications or combinations of medications. There were programs to improve follow up of other abnormal tests or identify patients who would benefit from vaccinations and health counseling.

This case study reviewed an example of how electronic health information tools can identify opportunities to improve population health. Although the original report did not discuss the actual impact on patient health outcomes, it described an overall framework for implementing such a program.


Danforth KN et al. Electronic Clinical Surveillance to Improve Outpatient Care: Diverse Applications Within an Integrated Delivery System. eGEMs (Generating Evidence & Methods to improve patient outcomes) 2014;2(1). Accessed on July 1, 2014.

Continue Reading