Exact and approximate probability distributions of evidence-based bundle composite compliance measures.
Several studies recently have indicated that compliance with the medical evidence base across a wide range of chronic and acute conditions is dismally low. Partly in response, core measure sets and evidence-based "bundles" of care processes known to improve a patient's condition have become a common...
| Publicado en: | Health Care Management Science Vol. 13; no. 3; pp. 193 - 210 |
|---|---|
| Autores principales: | , , , |
| Formato: | Journal Article |
| Publicado: |
Springer Nature
Sep2010
|
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=105080430&site=ehost-live header: @attributes: shortDbName: ccm uiTerm: 105080430 longDbName: CINAHL Complete uiTag: AN controlInfo: bkinfo: dissinfo: jinfo: jid: 13869620 BSE jtl: Health Care Management Science issn: 13869620 maglogo: N pubinfo: dt: Sep2010 vid: 13 iid: 3 pid: 237 pub: Springer Nature place: New York, New York artinfo: ui: 105080430 NLM20715304 2010751075 NLM20715304 105080430 ppf: 193 ppct: 17 formats: tig: atl: Exact and approximate probability distributions of evidence-based bundle composite compliance measures. aug: au: Benneyan JC Taseli A Benneyan, James C Taşeli, Aysun affil: New England Healthcare Engineering Center, Northeastern University, 360 Huntington Avenue, Boston, MA 02115, USA sug: subj: Medical Practice, Evidence-Based Treatment Errors Prevention and Control Safety Guideline Adherence Statistics and Numerical Data Models, Statistical ab: Several studies recently have indicated that compliance with the medical evidence base across a wide range of chronic and acute conditions is dismally low. Partly in response, core measure sets and evidence-based "bundles" of care processes known to improve a patient's condition have become a common means for assessing and comparing the reliability of healthcare organizations to comply with best care practices. Compliance typically is estimated with a composite reliability statistic of the percent of all indicated bundle elements actually delivered to a sample of patients, which then is used to monitor, improve, rank, and reward organizations. Given these judgments inherently involve statistical concepts, we derive exact probability distributions of bundle reliability measures and illustrate how they can be used to develop appropriate statistical methods (confidence intervals, hypothesis tests, statistical process control charts, and probability ratio tests) for supporting process improvement and patient safety activities. Due to their computational complexity, we also discuss several accurate and fast approximations that can be useful in practice, including cumulant-based expansions, saddle point approximations, and probability generating function inversion. In contrast, we show that in some cases simple binomial or normal approximations can result in significant interpretation error and prolonged delays to detect process improvements or deteriorations. pubtype: Academic Journal doctype: Journal Article ougenre: Article language: English refInfo: holdings: @attributes: islocal: N |
|---|