HIM 690 Module 3 Progress Report Example

Midpoint: A Progress Report on Data Collection for the Cimarron Heights Flag Accuracy Capstone

[Student Name]

Southern New Hampshire University

HIM 690: Health Information Management Capstone

Module Three Progress 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.

Midpoint: A Progress Report on Data Collection for the Cimarron Heights Flag Accuracy Capstone

This report covers the first six weeks of data collection for the Tulsa capstone that is comparing coded admission flags with a blinded second reading of each chart. It is written for the capstone advisor and the two sponsors, the hospital's nursing chief and its HIM director. It reports what has been completed, how reliable the work has been, what has gone wrong and what comes next.

Progress Against Plan

Two of three abstraction batches are finished. Of the 318 records in the sample, 214 have been abstracted, against a plan of 212 by the end of week six, so the count is on track. The breakdown by complication is 54 pressure wound records, 54 catheter infection records, 54 venous clot records and 52 fall injury records. Two records were excluded. In one, the patient was transferred from another hospital and the transfer documents were never scanned into the chart, so the condition at arrival could not be judged from the record at all. In the other, the coded condition had been entered on the wrong patient's account and corrected after billing, which made the original flag meaningless for the study. Both exclusions meet the exclusion rule for records missing essential documentation, and both were logged with reasons. One replacement record was drawn at random for the first; the second came from the fall stratum, which has no spares, so falls will finish at 77 records. The sample remains balanced enough across the four strata to describe each one separately as planned.

Reliability

Every tenth record in each batch was abstracted by both coders. Agreement in batch one was 31 of 33 records, kappa 0.86, and in batch two 32 of 34, kappa 0.88. McHugh (2012) places values from 0.80 to 0.90 in the strong range and cautions that percent agreement alone overstates reliability, which is why kappa is reported for each batch rather than a single overall figure. Both batches cleared the proposal's minimum of 0.80, so abstraction did not need to pause. The four disagreements were reviewed by the coding quality lead, and three were resolved by applying the written rules; the fourth led to the rule clarification described below.

A Rule Clarification

In batch two, the coders disagreed about a patient whose pressure wound was described in the emergency department note as a red area on the sacrum, then staged by the wound nurse on day two as a deep tissue injury. One coder judged the wound present on arrival; the other judged it clinically undetermined. The written rule covered deep tissue injuries first documented after day one but not a red area documented at arrival and staged later. With the advisor's approval, the rule was clarified: an area at the same site documented at arrival, even without a stage, counts as present on admission if the later staging confirms a pressure injury there. To avoid applying the clarification unevenly, both coders re-reviewed every pressure wound record already abstracted, 54 in all; the clarification changed the abstracted flag in three records, and those changes are logged.

Time and Workload

Abstraction is taking about 31 minutes per record against the 25 minutes planned. Almost all of the extra time goes to scanned outside records, such as nursing home transfer forms, which are stored as image files and must be read page by page. At the current pace, batch three will take one week longer than scheduled. The reserve week held in the charter will absorb this, so the overall schedule does not change, but there is no further buffer.

Two options for regaining time were considered and rejected. Skipping the scanned outside records would have been faster but would have biased the results, because those records are exactly where evidence of conditions present on arrival tends to sit for patients from facilities. Adding a third abstractor would have required new training and a new reliability check partway through the study. Absorbing the delay in the reserve week was the least risky choice.

Data Quality Checks

Following the verification approach described by Kahn et al. (2016), the writer runs three checks on the abstraction file each week: every required field is present, every coded value is one of the allowed choices and no assessment time falls before the admission time or after discharge. The checks found two blank fields and one impossible time in batch one, all corrected from the source records within a day, and none in batch two. A plausibility review compared the share of records with a documented 24-hour skin assessment across units. One unit's share, 92%, was well above the others, which range from 61% to 74%; a spot check confirmed that the unit uses a different admission template that requires the assessment, which is a finding worth carrying into the analysis rather than an error. Weiskopf and Weng (2013) recommend combining several assessment methods, and the weekly checks provide that alongside the double abstraction.

Early Observations

The study's analysis has not begun, and early counts should not be over-read. Still, two patterns are visible enough to shape the interview phase, which the charter allowed to begin early. Disagreement appears more frequent among pressure wound records than among the other three complications, and many disagreements involve patients admitted from long-term care facilities. Interviews will therefore include the two units admitting the most such patients.

These patterns will be tested, not assumed, once all records are in; if they fade in the full data, the interviews will still be useful for understanding how flags are set, but the report will say plainly that the early impression did not hold.

Next Steps and Request

Batch three begins Monday and should finish in two and a half weeks. Interviews start next week with four coders and four nurses. One request goes to the sponsors: the chief nursing officer is asked to release two night-shift nurses for 30-minute interviews during paid time, since recruitment has so far drawn only day-shift staff and the study's early patterns point to night admissions.

References

Kahn, M. G., Callahan, T. J., Barnard, J., Bauck, A. E., Brown, J., Davidson, B. N., Estiri, H., Goerg, C., Holve, E., Johnson, S. G., Liaw, S.-T., Hamilton-Lopez, M., Meeker, D., Ong, T. C., Ryan, P., Shang, N., Weiskopf, N. G., Weng, C., Zozus, M. N., & Schilling, L. (2016). A harmonized data quality assessment terminology and framework for the secondary use of electronic health record data. eGEMs, 4(1), Article 1244. https://doi.org/10.13063/2327-9214.1244

McHugh, M. L. (2012). Interrater reliability: The kappa statistic. Biochemia Medica, 22(3), 276-282. https://doi.org/10.11613/BM.2012.031

Weiskopf, N. G., & Weng, C. (2013). Methods and dimensions of electronic health record data quality assessment: Enabling reuse for clinical research. Journal of the American Medical Informatics Association, 20(1), 144-151. https://doi.org/10.1136/amiajnl-2011-000681