NURS 8211 Week 5 Chart Abstraction Tool Example

One Page, Between Admissions: A Chart Abstraction Tool and Protocol for Auditing Teach-Back Documentation on a Heart Failure Unit

Student Name

College of Nursing, Walden University

NURS 8211: Data Collection and Analysis for Evidence-Based Practice

Instructor Name

Month Day, Year

One Page, Between Admissions: A Chart Abstraction Tool and Protocol for Auditing Teach-Back Documentation on a Heart Failure Unit

Purpose and Design Constraints

The teach-back project on the composite heart failure unit needs a monthly audit of discharge charts to confirm what the navigator report shows and to capture information the report cannot, such as whether re-teaching was documented. Charge nurses will do the abstraction between admissions and other duties, which means the tool must be quick, unambiguous and tolerant of interruption. A tool that is precise but takes twenty minutes a chart will not be completed.

Retrospective chart review is often criticized for weak methods, including undefined variables, untrained abstractors and no check on agreement between reviewers (Gilbert et al., 1996). Guidance on chart review recommends a standardized form, explicit definitions and decision rules, abstractor training and a pilot to test the form before use (Vassar & Holzmann, 2013). This tool follows that guidance within the limits of a working unit. It has three parts: a one-page form, a protocol of decision rules and a record of the pilot that changed three of those rules.

Part 1: The Form

The form fits on one page and uses coded fields throughout, so that answers are marked, not written. Table 1 shows its fields.

Table 1

Teach-Back Chart Abstraction Form

Codes 9 (cannot tell) and 8 (not applicable) are kept separate from 0 (not recorded) so that missing documentation is not confused with a question that does not apply.

Part 2: Decision Rules

The protocol gives a rule for every case that abstractors found ambiguous in the pilot or that the project team expected. The main rules are:

Rule A. More than one discharge in the month. Abstract only the first eligible discharge for each patient in the month, so that frequent readmissions do not count twice.

Rule B. Response recorded in free text rather than the field. If the note states what the patient said about the rule, code it as if it were in the field. If the note says only that teaching was done, code 0.

Rule C. Caregiver answered instead of the patient. Code the response as given and mark field 7 as caregiver. Do not code the patient's own understanding as correct.

Rule D. Patient discharged against medical advice. Code field 2 as eligible if discharged home, and code fields 4 to 6 as documented; do not exclude these patients, since excluding them would hide a group at high risk.

Rule E. Interrupted abstraction. Write the field number reached on the form before leaving, and resume from that field. Do not start the chart again.

Rule F. Copied text. Mark field 9 as yes only if the teach-back wording matches an earlier note word for word, including the patient's quoted responses. Similar content in different words is not copied.

Part 3: The Pilot and the Three Rules It Tightened

The form was piloted on twelve charts from the month before relaunch. Two charge nurses abstracted the same charts independently, and their codes were compared field by field. Agreement was good on most fields but poor on three, and the disagreements traced to three rules that were too loose.

Field 4 to 6 codes. The first draft had only correct, not correct and not recorded. Abstractors disagreed on patients who answered wrongly and then correctly after re-teaching; one coded these correct and the other not correct. Code 2, correct after re-teaching, was added. This matters to the project, since re-teaching is part of teach-back, and a patient who needs it has still learned the rule.

Rule B. The first version said to code free-text responses as recorded. Abstractors disagreed on notes that said the patient verbalized understanding. The rule now requires the note to state what the patient said, since verbalized understanding describes the patient's agreement, not a teach-back.

Rule F. The first version asked abstractors to mark copied text if the note looked similar. Agreement was poor. The rule now requires word-for-word matching, which is slower but consistent.

After the changes, the two charge nurses abstracted six new charts and agreed on all fields. A formal agreement statistic such as kappa would be more informative with a larger sample (McHugh, 2012). The protocol therefore requires a second abstractor to code two charts each month, so that agreement can be tracked over time and the rules revisited if it drops.

Time and Feasibility

Field 10 records minutes per chart. In the pilot, the median time was about six minutes after abstractors had done three or four charts, which the unit council judged manageable for ten charts a month spread across four charge nurses. If abstraction time rises, the first field to drop would be field 9, which is the slowest and least central to the project's question.

Training

Each charge nurse will receive a thirty-minute session covering the form, the rules and two practice charts with answers. New abstractors will code two charts alongside an experienced abstractor before working alone. A laminated copy of the rules will be kept with the forms, since abstractors are likely to be interrupted and will need to look rules up quickly.

References

Gilbert, E. H., Lowenstein, S. R., Koziol-McLain, J., Barta, D. C., & Steiner, J. (1996). Chart reviews in emergency medicine research: Where are the methods? Annals of Emergency Medicine, 27(3), 305-308. https://doi.org/10.1016/S0196-0644(96)70264-0

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

Vassar, M., & Holzmann, M. (2013). The retrospective chart review: Important methodological considerations. Journal of Educational Evaluation for Health Professions, 10, Article 12. https://doi.org/10.3352/jeehp.2013.10.12