Nuts and bolts: An approach to improving clinical gaps in care

June 3, 2016 / By L. Gordon Moore, MD

I’m often asked by health system leadership “How do we improve X?” where X is a specific quality or utilization metric.  This blog is a description of one (not the only) way to improve a metric.  The approach is based on the science of improvement as well as aggregate experience of many engaged in quality improvement over the past two decades.

I’ve broken this down into three parts:  technology, process and people and use this framework to address a clinical gap in care (e.g. mammography rate or childhood immunization rate). 

I want to call out leadership and culture as critical components supporting quality improvement because of a wonderful study by Brewster et al.[1]  Hospitals that performed well on readmission reduction (according to their study) were different because they:

  • Supported collaboration across departments
  • Shared quality improvement data and expertise with community partners
  • Enthusiastically embraced trial-and-error
  • Made their patients the focus rather than the metrics or the finances

So in addition to following the good leadership and culture practices noted above, here are things a health system can do to improve gaps on care:


Work with your EMR/IT department to identify how the specific gap is recorded.  See if it is structured or unstructured data and if there are existing reports that can identify the individuals to whom the gap applies and can identify the appropriate time interval and gap.

Let’s take well-child checks in infants for example. The EMR is able to produce a report of all infants eligible for well child checks and flag those who have not had them according to the appropriate schedule. In addition, gaps in care reports are run on a regular basis and given to an outreach person and gaps in care are represented as alerts in the patient’s chart. The practice/group/system can run regular reports to identify rates of care gaps by patient, clinician, practice unit, practice, group, etc.


Adopt a standard approach and set of guidelines for prevention/chronic condition management (consider following the approach described as the Chronic Care Model) and create a regular meeting for clinical staff for review of guidelines, standards in care.

Next, adopt a model for improvement. There is reasonable science to guide process improvement and The Model for Improvement is a reasonable (but not the only) approach:

Develop an outreach function and role

  • Example: An individual is designated as the person who will contact the infant’s caregiver to recommend an in-person follow up for a well child check.  The outreach person receives a report on a regular basis (see Technology) identifying infants with missing well-child checks.

Develop the process for opportunistic gap closure

  • This is the process by which any member of the practice – when calling up a patient’s chart – will see an alert identifying a gap in care.
  • Any person with appropriate access to the patient’s record can see the care gap alert and recommend an appropriate response. Example: A mom stops by the practice to pick up a form for their child. The registrar sees that the child is overdue for a well-child check and recommends making an appointment.

Develop a process for standardizing data capture

  • EMR data tends to be unstructured – different clinicians record things differently. If a key data element is captured as text in a note “She had a normal screening mammogram last month.” – the data may be invisible to the reporting engines.  This means that the accuracy of reports may be low. 
  • Create a data standardization clinical meeting. Clinical groups working on improving outcomes often benefit from creating a regular venue for discussing how data are captured in their electronic systems so that they can better represent what is truly happening with their patients.


Outreach person
: The nature of the work defines the role.  If this person is reaching out to ask a person with a gap to make a follow-up appointment, then the role can be staffed by a non-clinical individual.  In this example if a patient or caregiver brings up medical issues, the non-clinical person would follow the appropriate process of handing that issue off to a clinician.

Analyst: Someone who understands the nature of the data and can regularly run reports.  Understanding the data means that they can understand the strengths and weaknesses of the data set from which the reports run.  This includes understanding delays in health plan data as well as lack of standardization in EMR data and how these factors impact the results.

Clinical champion/quality improvement director: A doctor or nurse with some experience in quality improvement should work closely with the report generator person to review the criteria used to define numerators and denominators.  This person should also at least participate in (if not lead) the data standardization meetings.

Process improvement person: It is often helpful to have someone with training in process improvement support these efforts in practice.  Consider six sigma training, quality improvement or quality advisor training.

Example of addressing a specific metric

Let’s take “low mammography rate.”  The six-step process below can fit any gap in care and does not make recommendations specific to mammography.

  1. A clinical group creates a clinical guidelines committee.
  2. The committee adopts a standard approach to mammography, e.g. the USPSTF mammography recommendations.
  3. A clinical quality improvement director and analyst work together to identify what data are available and how they might populate a numerator and denominator:
     – Denominator: Women eligible for mammography
     – Numerator: Eligible women who have not had mammography within the appropriate time interval
     – These data may be in the EMR, might be supplemented by health insurance claims or could be available in a health information exchange.
  4. QI director, analyst and process improvement person test and then implement a process for an alert that shows up in the patient’s chart and is addressed by any member of the care team.
    – One consideration is to create a standard care pathway where the default is to close the gap, e.g. “Our practice has a standing order to provide a flu shot for all eligible patients who meet the following criteria…”
  5. QI director, analyst and process improvement person create, test, and implement an outreach report and work with an outreach coordinator.
  6. The quality improvement team meets regularly to review results, problem solve and continually improve.

To reiterate: This is one approach to quality improvement that has led many organizations to better outcomes for the patients they serve.  I encourage you and your organization to experiment with this approach or one of the other methods out there. Find an approach that works for you.

L. Gordon Moore, MD, is senior medical director for Populations and Payment Solutions at 3M Health Information Systems.

[1] Brewster, Amanda L., Emily J. Cherlin, Chima D. Ndumele, Diane Collins, James F. Burgess, Martin P. Charns, Elizabeth H. Bradley, and Leslie A. Curry. “What Works in Readmissions Reduction: How Hospitals Improve Performance.” Medical Care 54, no. 6 (June 2016): 600–607. doi:10.1097/MLR.0000000000000530.