The use of error reporting data in patient safety research

Swana, Nelisiwe (2019-04)

Thesis (MMed)--Stellenbosch University, 2019.

Thesis

ENGLISH SUMMARY : Background: Error reporting is the only data source designed specifically to generate patient safety data, yet its ability to report comprehensively on patient safety has not been assessed. This study aimed to assess the capability of error reporting data to report comprehensively on adverse events. The research question that guided this study was: “What are the perceptions of patient safety experts regarding the use of error reporting data in patient safety?” The objectives of the study were set to determine the demographic characteristics of those who use error reporting in patient safety, i.e., their professional backgrounds, area of practice, countries, languages and years of experience with error reporting data. The study also aimed to determine how they perceive error reporting`s capability to report comprehensively on patient safety and assess whether their professions or areas of practice influenced their choice to work with error reporting and not with other data sources also used in patient safety. Methods: A cross-sectional survey using an online questionnaire was conducted. This was a descriptive study design with a quantitative approach. Reliability and validity of the study were assured through a pilot test and in consultation with the project leader (who was the researcher’s supervisor and an expert in epidemiology research), and experts in patient safety field at the University of Frankfurt. This study was conducted with approval from the University of Frankfurt (see annexure 1b) and University of Stellenbosch ethics committees. Ethics reference number: S15/11/267. Sixty-two experts in patient safety research were surveyed using a 15-item online questionnaire. Error reporting was measured against other data sources on its availability, accessibility, time consumption, annual report generation, ability to report on all causes and level of harm to patients, uniqueness, independence and adaptability to different health organisations. Data were analysed using Statistical Analysis Software 9.4. Results: Error reporting was the most widely used patient safety data source with 36 (58%) participants indicating a preference for using it, as opposed to 13 (21%) for chart review, 10 (16%) for claims data, 9 (15%) for routine data and 8 (13%) for survey data. Experts felt it was the best in reporting on all levels of harm (61.1%) and generating annual patient safety reports (44.4%), but it was the most inaccessible data source (37.2%). Both profession (p = 0.25) and area of practice (p = 0.79) had no influence on experts’ choice for error reporting as a data source. The difference between error reporting’s overall performance in patient safety and the other data sources, measured using a t-test was not statistically significant (between error reporting and claims data source (p = 0.92), between error reporting and routine data source (p = 0.74) and between error reporting and survey data source (p = 0.61). Conclusion: Error reporting, although the most widely used, had shortcomings in several areas currently complemented by other data sources. Thus, relying on it alone could be inadequate, and ways to integrate data sources should be explored further.

AFRIKAANSE OPSOMMING : Agtergrond: Fout-aanmelding is die enigste databron wat spesifiek vir die generasie van pasiënt veiligheid data ontwerp is, en tog is dié bron se vermoë om omvattend oor pasiënt veiligheid te rapporteer, nie geëvalueer nie. Die doel van die studie was om te beoordeel of die fout-aanmelding data volledig kan rapporteer op ongunstige gebeure, aangesien dit die enigste databron ontwerp is vir pasiënt veiligheid. Die navorsing vraag wat die studie gelei het was: “Wat was die persepsie van pasiënt veiligheid kenners met betrekking tot die verteenwoordiging en evaluasie van fout-aanmelding data in pasiënt veiligheid?” Die objektief van die studie was om die demografiese eienskappe te bepaal van die wat fout-aanmelding in pasiënt veiligheid gebruik. Hulle professionele agtergrond, praktyk-gebied, lande, tale en jare van ondervinding wat fout-aanmelding data rapporteur bepaal hoe hulle fout-aanmelding sien en die vermoë om dit te rapporteur volledig op pasiënt veiligheid en dit beoordeel of hulle beroep praktyk beïnvloed die keuse of hulle fout-aanmelding en nie ander databronne gebruik in pasiënt veiligheid. Metodes: In ’n opname oor pasiënt veiligheid was 62 deskundiges geraadpleeg en ’n aanlyn vraelys bestaande uit 15 items gebruik. Fout-aanmelding in verhouding tot ander databronne was gemeet, met inagneming van die beskikbaarheid, toeganklikheid, tydverloop, jaarlikse verslag generasie, vermoë om oor alle oorsake en vlakke van pasiënt bedreiging te rapporteer, uniekheid, onafhanklikheid, asook die aanpasbaarheid daarvan by verskillende organisasies. Data is met behulp van die rekenaarsagteware Statistical Analysis Software 9.4 ontleed. Resultate: Fout-aanmelding was die algemeenste pasiënt veiligheid databron by 36 (58%) van die deelnemers, in vergelyking met 13 (21%) se voorkeur aan kaart hersiening, 10 (16%) se voorkeur aan eise data, 9 (15%) se voorkeur aan roetine data en 8 (13%) se voorkeur aan opname data. Kundiges was van mening dat foutopsporing die beste presteer het wat rapportering op alle vlakke van bedreiging (61.1%) betref, asook vir die skep van jaarlikse pasiënt veiligheid verslae (44.4%), maar dat fout-aanmelding die ontoeganklikste databron (37.2%) was. Daar is bevind dat sowel beroep (p = 0.25) as praktyk gebied (p = 0.79) geen invloed op die deskundiges se keuse van fout-aanmelding as ’n databron gehad het nie. Volgens ’n toets wat fout-aanmelding en ander bronne se oorhoofse prestasie ten opsigte van pasiënt veiligheid gemeet het, was die verskil statisties onbeduidend (tussen fout-aanmelding en eis databron (p = 0.92), tussen fout-aanmelding en roetine databron (p = 0.74), en tussen fout-aanmelding en opname databron (p = 0.61)). Gevolgtrekking: Alhoewel fout-aanmelding die algemeenste gebruik is, het dit tekortkominge op verskillende gebiede getoon, wat tans deur ander databronne aangevul word. Deur alleenlik op fout-aanmelding staat te maak kan ontoereikend wees, en maniere om databronne te integreer behoort gevolglik verder verken te word.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/106243
This item appears in the following collections: