How Valid is the AHRQ Patient Safety Indicator "Postoperative Respiratory Failure"?
How Valid is the AHRQ Patient Safety Indicator "Postoperative Respiratory Failure"?
复制标题
DOI：
10.1016/j.jamcollsurg.2010.09.034
复制
发表时间：
2011-06-01
影响因子：
5.2
通讯作者：
Rosen, Amy K.
中科院分区：
文献类型：
--
作者：
Borzecki, Ann M.;Kaafarani, Haytham M. A.;Rosen, Amy K.
背景：医疗研究和质量患者安全指标机构术后呼吸衰竭(PRF)使用行政数据，根据呼吸衰竭诊断或插管或呼吸程序代码筛查选择性手术后潜在的可预防的呼吸衰竭。关于PRF在识别真实事件方面的准确性的数据很少；最近使用大学卫生系统联盟数据进行的一项研究发现，阳性预测值(PPV)为83%。我们在退伍军人健康管理局中检查了该指标的PPV。研究设计：我们将患者安全指标软件(v.3.1a)应用于2003-2007财年退伍军人管理局的VA出院数据。训练有素的抽象员审查了112个带有软件标记的PRF病例的医疗记录。结果：75例为真阳性(PPV为67%；95%CI为57-76%)；13%为诊断代码，53%为程序代码，两者均为33%。在假阳性中，19%代表编码错误，76%代表非选择性入学。在真正阳性的患者中，28%的患者死亡，56%的患者美国麻醉师协会的水平高于II级。在相关的指标性操作中，53%的患者死于腹部/盆腔，56%的患者持续3小时。结论：根据我们和大学卫生系统联合会的研究结果，PRF应该继续用作潜在患者安全事件的筛查。它的PPV可以通过引入入院状态代码在退伍军人健康管理局得到实质性改善。根据患者和手术相关因素，许多经PRF确诊的病例似乎处于高风险状态。这种情况在多大程度上是真正可以预防的事件，需要进行额外的评估。(J am Coll Surg 2011；212：935-945。(C)2011年由美国外科医师学会颁发)
BACKGROUND: The Agency for Healthcare Research and Quality Patient Safety Indicator postoperative respiratory failure (PRF) uses administrative data to screen for potentially preventable respiratory failure after elective surgery based on a respiratory failure diagnosis or an intubation or ventilation procedure code. Data on PRF accuracy in identifying true events is scant; a recent study using University HealthSystem Consortium data found a positive predictive value (PPV) of 83%. We examined the indicator's PPV in the Veterans Health Administration.STUDY DESIGN: We applied the Patient Safety Indicator software (v.3.1a) to fiscal year 2003-2007 VA discharge data. Trained abstractors reviewed medical records of 112 software-flagged PRF cases. We calculated the PPV and examined false positives to determine reasons for incorrect identification and true positives to determine clinical consequences and potential risk factors of PRF.RESULTS: Seventy-five cases were true positive (PPV 67%; 95% CI, 57-76%); 13% were identified by a diagnosis code, 53% by a procedure code, 33% by both. Of false positives, 19% represented coding errors, 76% represented nonelective admissions. Of true positives, 28% of patients died, 56% had an American Society of Anesthesiologists level higher than II. Of associated index procedures, 53% were abdominal/pelvic, and 56% lasted >3 hours.CONCLUSIONS: Based on our and University HealthSystem Consortium's findings, PRF should continue to be used as a screen for potential patient-safety events. Its PPV could be substantially improved in the Veterans Health Administration through introduction of an admission status code. Many PRF-identified cases appeared to be at high risk, based on patient and procedure-related factors. The degree to which such cases are truly preventable events requires additional assessment. (J Am Coll Surg 2011;212:935-945. (C) 2011 by the American College of Surgeons)