HIM 480 Module 7 Final Capstone Project Example

Final Capstone Project: Making the Problem List Tell the Truth, Measuring and Improving Problem List Accuracy at Hollis Ridge Health

[Student Name]

Southern New Hampshire University

HIM 480: Health Information Management Capstone

Final Capstone Report

[Instructor Name]

[Date]

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

Final Capstone Project: Making the Problem List Tell the Truth, Measuring and Improving Problem List Accuracy at Hollis Ridge Health

Abstract

Electronic problem lists drive decision support, quality measurement, risk adjustment and the health summaries patients read, yet their accuracy is rarely checked. This capstone audited 400 randomly selected adult primary care patients at Hollis Ridge Health against condition-specific rules for six chronic conditions and interviewed 12 clinicians. Of 569 qualifying conditions, 72.6% were listed; completeness ranged from 88% for hypertension to 41% for chronic kidney disease. Resolved problems remained active on 23% of lists and duplicates appeared on 14%. Root causes included unclear ownership, no routine review, a cumbersome interface, missing prompts and no policy. A six-part program and a staged evaluation plan are recommended.

Background

The problem-oriented record, which Weed (1968) proposed to organize care around each of a patient's problems, gave the problem list its central place. Modern records have multiplied its uses. At Hollis Ridge, a drug safety rule failed to warn a prescriber because a patient's heart failure was absent from his list, which prompted this project. Before it, no one at the system knew how often lists were incomplete or out of date.

What the Research Shows

Published studies agree that problem lists are frequently incomplete and that completeness varies across organizations and conditions; Wright et al. (2015) documented that variation across ten sites and linked better results to policy, culture and design. Holmes et al. (2012) showed that clinicians disagree about who maintains the list and what belongs on it. Hodge and Narus (2018) identified ownership, usability, standard terms and policy as recurring success factors. The strongest intervention evidence comes from a randomized trial by Wright et al. (2012), in which suggestions of likely missing problems, inferred from medications and laboratory results, increased documentation of targeted conditions. Few studies measured whether resolved problems are kept current, and most came from large academic centers.

Methods

The sampling frame was every adult seen twice or more in primary care over the prior year; 400 were drawn at random. For each of the six target conditions (hypertension, depression, COPD, kidney disease, heart failure and type 2 diabetes), written rules set out when the condition belonged on the list, using documented diagnoses, medications and laboratory values. Reviewers recorded listing, codability, stale entries and duplicates; two reviewers double-checked 40 charts and agreed on 94% of judgments. Twelve clinicians were interviewed, and the screens used to maintain the list were reviewed. The system classified the work as quality improvement.

Results

Table 1 summarizes the main findings. Completeness was lowest for chronic kidney disease and depression and ranged from 81% to 61% across clinics, lower where many part-time clinicians worked. Only 6% of listed entries were free text.

Table 1. Key Results

Note. Baseline audit of adult primary care patients; composite data.

Why the Gaps Exist

Clinician interviews and the record review pointed to five causes. No one felt responsible for the whole list, so primary care and specialists each expected the other to update it. There was no moment in any visit when the list was reviewed. Adding or resolving a problem required leaving the note. The record never suggested a condition when laboratory data implied one, which explains why declining kidney function so often went unlisted. And no policy defined ownership or content. Clinicians also questioned whether the list was worth maintaining after years of seeing it cluttered, a cycle in which poor quality discourages the use that would improve it.

The Recommended Program

The program has six parts, each tied to a cause. A medical staff policy assigns overall ownership to the primary care clinician and defines what belongs on the list. Rooming nurses review the list with patients at chronic care visits, and clinicians confirm or resolve flagged items. A problem pane inside the note template makes updates part of documentation. Decision support suggests likely missing kidney disease, diabetes and heart failure, displayed only in the pane, never as interrupting alerts. Clinics receive monthly reports with an example of how the list is used. Patients can flag outdated problems in the portal for the care team to review. The health information team removes duplicates and maintains approved terms.

Evaluation Plan

Three clinics, including the weakest, will pilot the program for 12 weeks while the other nine continue as before; the nine will then adopt it in monthly waves. Suggestion rules will run silently for two weeks before clinicians see them. Targets for twelve months are 90% combined completeness, 80% for chronic kidney disease and no more than 8% of lists with stale problems, with balancing measures for visit length, dismissed suggestions and clinician satisfaction. Quarterly audits using the same rules will track outcomes, and process measures will be charted monthly. Cresswell et al. (2013) stress continued support after launch in health IT change, so champions and monthly reports continue beyond the pilot.

Discussion

The findings matter beyond tidiness. When a quarter of qualifying conditions are missing, drug safety and dosing rules fall silent for those patients, quality measures undercount the population, care managers miss people who need outreach and risk-adjusted payments understate how sick patients are. The gaps were not random: they concentrated in conditions detected through laboratory trends and in clinics with fragmented care, which suggests that prompts and clear ownership will help most where they are needed most. The currency finding adds to a literature that has focused mostly on missing conditions.

Limitations

Rules based on structured data may have missed conditions described only in narrative notes. The project covered one system's primary care clinics, and interviewed clinicians may have been more engaged than others. Implementation and outcomes fall outside the course timeline, so the program's effect remains to be demonstrated.

Recommendations

The sponsor is asked to approve four actions: submit the ownership policy to the medical executive committee, fund 60 informatics hours to build the problem pane and suggestion rules, launch the three-clinic pilot next quarter and assign the quarterly audit and monthly reports to the health information data integrity team as a standing duty.

Conclusion

Hollis Ridge's problem lists are incomplete in predictable ways for predictable reasons. By giving the list an owner, a routine, an easier path, timely prompts, feedback and a patient check, the recommended program aims to make the list tell the truth about each patient, which is what every system that reads it assumes.

References

Cresswell, K. M., Bates, D. W., & Sheikh, A. (2013). Ten key considerations for the successful implementation and adoption of large-scale health information technology. Journal of the American Medical Informatics Association, 20(e1), e9-e13. https://doi.org/10.1136/amiajnl-2013-001684

Hodge, C. M., & Narus, S. P. (2018). Electronic problem lists: A thematic analysis of a systematic literature review to identify aspects critical to success. Journal of the American Medical Informatics Association, 25(5), 603-613. https://doi.org/10.1093/jamia/ocy011

Holmes, C., Brown, M., Hilaire, D. S., & Wright, A. (2012). Healthcare provider attitudes towards the problem list in an electronic health record: A mixed-methods qualitative study. BMC Medical Informatics and Decision Making, 12, Article 127. https://doi.org/10.1186/1472-6947-12-127

Weed, L. L. (1968). Medical records that guide and teach. New England Journal of Medicine, 278(11), 593-600. https://doi.org/10.1056/NEJM196803142781105

Wright, A., McCoy, A. B., Hickman, T.-T. T., Hilaire, D. S., Borbolla, D., Bowes, W. A., Dixon, W. G., Dorr, D. A., Krall, M., Malhotra, S., Bates, D. W., & Sittig, D. F. (2015). Problem list completeness in electronic health records: A multi-site study and assessment of success factors. International Journal of Medical Informatics, 84(10), 784-790. https://doi.org/10.1016/j.ijmedinf.2015.06.011

Wright, A., Pang, J., Feblowitz, J. C., Maloney, F. L., Wilcox, A. R., McLoughlin, K. S., Ramelson, H., Schneider, L., & Bates, D. W. (2012). Improving completeness of electronic problem lists through clinical decision support: A randomized, controlled trial. Journal of the American Medical Informatics Association, 19(4), 555-561. https://doi.org/10.1136/amiajnl-2011-000521