145 - Can education lead to a sustained improvement in neonatology clinical documentation integrity?

Poster Number: 145
Publication Number: 145.218

Ashish O. Gupta, Nemours Children's Hospital, Wilmington, DE, United States

Background: The current medical education curriculum does not include a formal education on clinical documentation strategies, even though the physician documentation directly impacts the medical decision making, and ultimately the revenues. In an external audit, we identified a substantial gap in the clinical documentation for medical decision making, diagnosis selection and risk stratification, which is an essential requirement for optimal billing or DRG code identification. Physician education and an effective strategy to improve documentation is paramount to bridge this gap.

Objective: Increase the accuracy of clinical documentation for common Neonatology diagnoses by 80 % within 1 year.

Design/Methods: A quality improvement project was performed at a 20-bed level III NICU from May 2020-Dec 2021. Retrospective data was collected over 6 months from May-Oct 2020 for infants born at < 34 weeks of gestational age- clinical documentation and diagnosis accuracy for 5 common neonatology diagnoses: respiratory distress, apnea of prematurity, neonatal sepsis, jaundice and feeding immaturity. A scoring tool was developed and % accuracy was collected for clinical documentation and diagnosis selection for each of the five diagnoses. An educational tool was developed for providers, EMR optimization (key documentation phrases and reorganizing the preferred diagnosis list) performed, and education provided to billing and coding team. A periodic review performed by coding integrity team (2-3 times/week) and project champion (weekly)- queries generated in EMR and feedback provided to clinicians.

Results: A total of 150 patient charts were reviewed during the study period. Retrospective data collection revealed the average accuracy of clinical documentation ~42%, accuracy for selection of correct diagnosis was 27% and overall accuracy was ~36%. Among individual diagnoses, accuracy of documentation and diagnosis code selection was worse for respiratory distress ( < 20%), and best for jaundice (~50%) (Table 1). After educational intervention, accuracy of clinical documentation and diagnosis selection both improved for all 5 categories significantly and sustained for more than one year (Fig 1. A slight dip in accuracy noted in July-Aug 2021, which was most likely secondary to addition of new providers.Conclusion(s): A multidisciplinary comprehensive educational approach could lead to significant improvement in clinical documentation and selection of billing codes for common neonatology diagnoses. Our future plan is to review the financial impact of improved clinical documentation and medical decision making.
Monthly overall accuracy of clinical documentation and selection of diagnosis for all five categoriesMonthly overall accuracy of clinical documentation and selection of diagnosis for all five categories
Monthly run chartsFigure 1 a: Overall accuracy of clinical documentation and diagnosis
Figure 1 b: Accuracy of clinical documentation
Figure 1 c: Accuracy of selection of disgnosis