Physician derived versus administrative data in identifying surgical complications. Fact versus Fiction

Henry, Leonard R.; Minarich, Michael J.; Griffin, Rhonda; Holzen, Urs von; Hardy, Ashley; Fornalik, Hubert; Schwarz, Roderich E.

doi:10.1016/j.amjsurg.2018.08.015

Cited by 5 publications

(6 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…36 Similar to another report, 37 this previous study 36 found large positive likelihood ratios with moderate sensitivity. Another study 38 Funding/Support: The study was supported by the Big Ideas Program, a BJC HealthCare and Washington University School of Medicine internal grant program hosted by the Healthcare Innovation Lab and the Institute for Informatics (Dr Abraham). The data used for this study were in part supported by grant 1622678 from the National Science Foundation and grant R21 HS24581-01 from the Agency for Healthcare Research and Quality (Dr Avidan).…”

Section: Limitationsmentioning

confidence: 99%

“…36 Similar to another report, 37 this previous study 36 found large positive likelihood ratios with moderate sensitivity. Another study 38 found that practitioners and coders have substantial disagreement, largely around the severity of a complication. Others 39,40 Funding/Support: The study was supported by the Big Ideas Program, a BJC HealthCare and Washington University School of Medicine internal grant program hosted by the Healthcare Innovation Lab and the Institute for Informatics (Dr Abraham).…”

Section: Limitationsmentioning

confidence: 99%

See 1 more Smart Citation

Use of Machine Learning to Develop and Evaluate Models Using Preoperative and Intraoperative Data to Identify Risks of Postoperative Complications

et al. 2021

View full text Add to dashboard Cite

Key Points Question Can machine learning models predict patient risks of postoperative complications related to pneumonia, acute kidney injury, deep vein thrombosis, delirium, and pulmonary embolism? Findings In a cohort study of 111 888 operations at a large academic medical center, machine learning algorithms exhibited high areas under the receiver operating characteristic curve for predicting the risk of postoperative complications related to pneumonia, acute kidney injury, deep vein thrombosis, pulmonary embolism, and delirium. Meaning These findings suggest that machine learning models using preoperative and intraoperative data can predict postoperative complications and generate reliable and clinically meaningful interpretations for supporting clinical decisions along the perioperative care continuum.

show abstract

Section: Limitationsmentioning

confidence: 99%

Section: Limitationsmentioning

confidence: 99%

Use of Machine Learning to Develop and Evaluate Models Using Preoperative and Intraoperative Data to Identify Risks of Postoperative Complications

et al. 2021

View full text Add to dashboard Cite

show abstract

“…In the Netherlands, physicians register diagnostic codes thus avoiding remote coding by dedicated coders based on discharge summaries, which may introduce inaccuracies. [25][26][27] Our study has limitations. Only prescribed medications are recorded, thus excluding over the counter drugs.…”

Section: Discussionmentioning

confidence: 94%

Medication burden in epilepsy: Exploring the impact of non-epilepsy concomitant drugs load

Bunschoten

Palen

Sander

et al. 2020

Seizure

View full text Add to dashboard Cite

To determine the burden of non-epilepsy drugs on people with epilepsy, using administrative health care data. Methods:The Achmea Health Insurance Database (AHID) contains health claims data from 25% of the Dutch population. From the AHID, we selected all policyholders with coverage for at least one full calendar year between 2006-2009. We included adults with diagnostic codes for epilepsy and randomly selected two frequency-matched controls per case. We labeled drugs dispensed at least twice per calendar year as chronic and excluded antiseizure medications. We estimated and compared the prevalence of chronic medication use, number of chronic medications used, number of prescriptions dispensed, Rx Risk comorbidity index, and drug burden index (DBI) between people with epilepsy and controls.Results: Non-epilepsy chronic medication use was more frequent in people with epilepsy than controls (67% versus 59%, p<0.001). People with epilepsy had an increased DBI (average 0.19 versus 0.10, p<0.001), used more chronic medications (median 2 versus 1, p<0.001) and had more prescriptions dispensed (median 7 versus 3, p<0.001). The DBI and number of unique chronic medications were higher among older (>60 years) than younger (<60 years) subjects in cases and controls. Non-epilepsy chronic medication use was more prevalent in people with epilepsy across all therapeutic drug classes and most comorbidities measured using the Rx Risk score. Conclusion:Chronic non-epilepsy medication use is more prevalent among people with epilepsy. The medication burden is higher among elderly with epilepsy and could partially explain the lower quality of life of people with epilepsy with comorbidities.

show abstract

“…15 Approximately 23 per cent of American surgeons now participate in some fashion in accountable care organizations; however, their role in a process that is more focused on comprehensive, cost-efficient delivery of chronic care is still not well understood. 2,11,[16][17][18][19][20] What is apparent is that acute surgical care must be approached as a coordinated interaction of multiple specialists who are committed to understanding and implementation of a patient-centered culture of quality. 1,9,10,16,[19][20][21][22][23] Our previous analysis of the impact of comorbid conditions on surgical outcome demonstrated results that were equal or better than those listed for the NSQIP national sample.…”

Section: Discussionmentioning

confidence: 99%

“…The final arbiter of surgical quality must be the surgeon and must be defined in specific and objective terms. [9][10][11][12][13] We hypothesized that participation in NSQIP would be associated with measurable improvement in surgical outcomes as indicated by incidence and effect of postoperative AEs over time. By supplementing the NSQIP quality measurement program with the C-D adverse event classification system, the burden of unanticipated AEs can be measured in objective terms, thereby defining quality as progression toward "zero defects" care.…”

mentioning

confidence: 99%

Clavien-Dindo Analysis of NSQIP Data Objectively Measures Patient-Focused Quality

Gurien

Crandall

et al. 2019

The American Surgeon

View full text Add to dashboard Cite

Current quality measures intended to drive improved clinical performance are perceived as an inappropriate administrative burden. Surgeon-constructed quality measures, including the NSQIP, are more closely aligned with provider performance and relevant outcome. We hypothesized that NSQIP participation would be associated with measurable improvement in surgical outcomes. Elective general surgical cases were compared by case volume and incidence of postoperative adverse events (AEs) from 2014 to 2017. Using the Clavien-Dindo severity scaling system, we summed the grades for each AE and defined the patient population burden of these AEs as this sum divided by case volume. Case volume samples increased 67 per cent from 2014 (n = 526, 30 day complete) to 2017 (n = 878). Ratio of patient burden to case volume improved from 0.92 (2014) to 0.73 (2017). Comparison of AE incidence was not significantly different; however, the majority decreased over time. Analysis of individual AE interval change identified sepsis-related respiratory care as the top priority performance improvement target. These data reflect improved performance for a growing volume of surgical procedures. The impact of perioperative morbidity and their associated burden on affected patients has decreased, demonstrating the value of combining NSQIP with Clavien-Dindo to measure the quality of surgical care in objective and patient-specific terms.

show abstract

Physician derived versus administrative data in identifying surgical complications. Fact versus Fiction

Cited by 5 publications

References 7 publications

Use of Machine Learning to Develop and Evaluate Models Using Preoperative and Intraoperative Data to Identify Risks of Postoperative Complications

Use of Machine Learning to Develop and Evaluate Models Using Preoperative and Intraoperative Data to Identify Risks of Postoperative Complications

Medication burden in epilepsy: Exploring the impact of non-epilepsy concomitant drugs load

Clavien-Dindo Analysis of NSQIP Data Objectively Measures Patient-Focused Quality

Contact Info

Product

Resources

About