Dataset Information

Developing Machine Learning Algorithms to Support Patient-centered, Value-based Carpal Tunnel Decompression Surgery.

ABSTRACT:

Background

Carpal tunnel syndrome (CTS) is extremely common and typically treated with carpal tunnel decompression (CTD). Although generally an effective treatment, up to 25% of patients do not experience meaningful benefit. Given the prevalence, this amounts to considerable morbidity and cost without return. Being able to reliably predict which patients would benefit from CTD preoperatively would support more patient-centered and value-based care.

Methods

We used registry data from 1916 consecutive patients undergoing CTD for CTS at a regional hand center between 2010 and 2019. Improvement was defined as change exceeding the respective QuickDASH subscale's minimal important change estimate. Predictors included a range of clinical, demographic and patient-reported variables. Data were split into training (75%) and test (25%) sets. A range of machine learning algorithms was developed using the training data and evaluated with the test data. We also used a machine learning technique called chi-squared automatic interaction detection to develop flowcharts that could help clinicians and patients to understand the chances of a patient improving with surgery.

Results

The top performing models predicted functional and symptomatic improvement with accuracies of 0.718 (95% confidence interval 0.660, 0.771) and 0.759 (95% confidence interval 0.708, 0.810), respectively. The chi-squared automatic interaction detection flowcharts could provide valuable clinical insights from as little as two preoperative questions.

Conclusions

Patient-reported outcome measures and machine learning can support patient-centered and value-based healthcare. Our algorithms can be used for expectation management and to rationalize treatment risks and costs associated with CTD.

SUBMITTER: Harrison CJ

PROVIDER: S-EPMC9015194 | biostudies-literature | 2022 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Developing Machine Learning Algorithms to Support Patient-centered, Value-based Carpal Tunnel Decompression Surgery.

Harrison Conrad J CJ Geoghegan Luke L Sidey-Gibbons Chris J CJ Stirling Paul H C PHC McEachan Jane E JE Rodrigues Jeremy N JN

Plastic and reconstructive surgery. Global open 20220418 4

<h4>Background</h4>Carpal tunnel syndrome (CTS) is extremely common and typically treated with carpal tunnel decompression (CTD). Although generally an effective treatment, up to 25% of patients do not experience meaningful benefit. Given the prevalence, this amounts to considerable morbidity and cost without return. Being able to reliably predict which patients would benefit from CTD preoperatively would support more patient-centered and value-based care.<h4>Methods</h4>We used registry data fr ...[more]

PMID: 35450263

Similar Datasets

Project description:BackgroundCarpal tunnel decompression surgery to treat carpal tunnel syndrome is a common procedure, yet data on safety and effectiveness of the operation in the general population remain scarce. We aimed to estimate the incidence of reoperation and serious postoperative complications (requiring admission to hospital or further surgery) following carpal tunnel decompression in routine clinical practice and to identify the patient factors associated with these adverse outcomes.MethodsWe did a nationwide cohort analysis including all carpal tunnel decompression surgeries in patients aged 18 years or older, done in the National Health Service in England between April 1, 1998, and March 31, 2017, using the Hospital Episode Statistics dataset linked to mortality records. Patients were followed-up until death or until the end of the study (March 31, 2017). Primary outcomes were the overall incidence of carpal tunnel decompression reoperation and serious postoperative complications (surgical site infection or dehiscence, or neurovascular or tendon injury, requiring admission to hospital or further surgery) within 30 days and 90 days after surgery. Multivariable Cox regression analysis was used to identify factors influencing complications and reoperation, and the Fine and Gray method was used to adjust for the competing risk of mortality. This study is registered with ClinicalTrials.gov, NCT03573765.Findings855 832 carpal tunnel decompression surgeries were done between April 1, 1998, and March 31, 2017 (incidence rate 1·10 per 1000 person-years [95% CI 1·02-1·17]). 29 288 procedures (3·42%) led to carpal tunnel decompression reoperation (incidence rate 3·18 per 1000 person-years [95% CI 3·12-3·23]). Of the 855 832 initial surgeries, 620 procedures (0·070% [95% CI 0·067-0·078]) led to a serious complication within 30 days after surgery, and 698 procedures (0·082% [0·076-0·088]) within 90 days. Local complications within 90 days after surgery were associated with male sex (adjusted hazard ratio 2·32 [95% CI 1·74-3·09]) and age category 18-29 years (2·25 [1·10-4·62]). Male sex (adjusted subhazard ratio 1·09 [95% CI 1·06-1·13]), old age (>80 years vs 50-59 years: 1·09 [1·03-1·15]), and greater levels of comorbidity (Charlson score ≥5 vs 0: 1·25 [1·19-1·32]) and socioeconomic deprivation (most deprived 10% vs least deprived 10%: 1·18 [1·10-1·27]) were associated with increased reoperation risk.InterpretationTo our knowledge, this is the largest national study on carpal tunnel decompression to date, providing strong evidence on serious postoperative complication and reoperation rates. Carpal tunnel decompression appears to be a safe operation in most patients, with an overall serious complication rate (requiring admission to hospital or further surgery) of less than 0·1%.FundingVersus Arthritis; Medical Research Council; Royal College of Surgeons of England and National Joint Registry research fellowship; University of Oxford; National Institute for Health Research; and National Institute for Health Research Biomedical Research Centre, Oxford.

Project description:ImportanceCarpal tunnel release (CTR) technique may influence the likelihood of revision surgery. Prior studies of revision CTR following endoscopic CTR (ECTR) compared with open CTR (OCTR) have been limited by sample size and duration of follow-up.ObjectiveTo estimate the incidence of revision CTR following ECTR compared with OCTR in a national cohort.Design, setting, and participantsThis retrospective cohort study used data from the US Veterans Health Administration. Participants included all adults (age ≥18 years) undergoing at least 1 outpatient CTR from October 1, 1999, to May 20, 2021. Data were analyzed from May 21, 2021, to November 27, 2023.ExposureIndex CTR technique.Main outcomes and measuresThe primary outcome was time to revision CTR, defined as repeat ipsilateral CTR during the study period. Secondary outcomes were indications for revision, findings during revision, and additional procedures performed during revision.ResultsAmong 134 851 wrists from 103 455 patients (92 510 [89.4%] male; median [IQR] age, 62 [53-70] years) undergoing at least 1 CTR, 1809 wrists underwent at least 1 revision at a median (IQR) of 2.5 (1.0-3.8) years. In competing-risks analysis, the cumulative incidence of revision was 1.06% (95% CI, 0.99%-1.12%) at 5 years and 1.59% (95% CI, 1.51%-1.67%) at 10 years. ECTR was associated with increased hazard of revision CTR compared with OCTR (adjusted hazard ratio [aHR], 1.56; 95% CI, 1.34-1.81; P < .001). The risk difference for revision CTR associated with ECTR compared with OCTR was 0.57% (95% CI, 0.31%-0.84%) at 5 years (number needed to harm, 176) and 0.72% (95% CI, 0.36%-1.07%) at 10 years (number needed to harm, 139). Regardless of index CTR technique, the most common indication for revision was symptom recurrence (1062 wrists [58.7%]). A reconstituted transverse carpal ligament (TCL) was more common after ECTR compared with OCTR, whereas scarring of the overlying tissues and of the median nerve itself were more common following OCTR. Incomplete transverse-carpal-ligament release was observed in 251 of the wrists undergoing revision CTR (13.94%) and was more common among revisions following ECTR (odds ratio, 1.62; 95% CI, 1.11-2.37; P = .01).Conclusions and relevanceIn this cohort study of revision CTR in the Veterans Health Administration, ECTR was associated with increased risk of revision compared with OCTR, but the absolute risk was low regardless of technique. Intraoperative findings at revision varied significantly according to index CTR technique.

Project description:Background: The present study analyzes the effects of preoperative serum albumin, hematocrit, and creatinine on postoperative outcomes in patients undergoing carpal tunnel decompression surgery. Methods: The American College of Surgeons National Quality Improvement Program (NSQIP) database was queried from 2011 to 2020. Albumin, hematocrit, and creatinine were collected for each patient, alongside covariates. Outcomes included 30-day medical complications, 30-day wound complications, return to the operating room, nonhome discharge, and extended postoperative length of stay. Bivariate t-tests and multivariate logistic regressions were conducted. For any outcome-laboratory value pairs with significance on regression, area under the receiver operating characteristic curves (AUC) were constructed. Results: A total of 1440 patients with albumin, 3138 patients with hematocrit, and 3159 patients with creatinine levels were identified. Increased serum albumin was associated with lower odds of medical complications (aOR: 0.479, p=0.035). An overall cohort cutoff of ≤ 3.5 g/dL (AUC: 0.79, p < 0.001) was predictive of medical complications. On multivariate logistic regression, increased hematocrit reduced the odds of medical complications (aOR: 0.889, p < 0.001). Predictive hematocrit cutoffs of ≤ 39.7% (AUC: 0.77, p < 0.001) and ≤ 36.6% (AUC: 0.74, p < 0.001) were identified for medical complications amongst male and female patients, respectively. Similarly, increased serum creatinine was associated with greater odds of medical complications (aOR: 1.684, p=0.006). Creatinine cutoffs of ≥ 1.2 mg/dL (AUC: 0.58, p=0.033) and ≥ 1.0 mg/dL (AUC: 0.59, p=0.039) were identified for medical complications amongst male and female patients, respectively. Conclusions: Multiple preoperative serum values were predictive of postoperative medical complications, and laboratory value thresholds were identified in this carpal tunnel decompression cohort to aid in risk stratification.

Project description:BackgroundThe American Society of Anesthesiologists (ASA) Choosing Wisely Top-5 list of activities to avoid includes "Don't obtain baseline laboratory studies in patients without significant systemic disease (ASA I or II) undergoing low-risk surgery - specifically complete blood count, basic or comprehensive metabolic panel, coagulation studies when blood loss (or fluid shifts) is/are expected to be minimal." Accordingly, we define low-value preoperative tests (LVTs) as those performed before minor surgery in patients without significant systemic disease. The objective of the current study was to examine the extent, variability, drivers, and costs of LVTs before carpal tunnel release (CTR) surgeries in the US Veterans Health Administration (VHA).MethodsUsing fiscal year (FY) 2015-2017 data derived from the VHA Corporate Data Warehouse (CDW), we determined the overall national and facility-level rates and associated costs of receiving any of 8 common LVTs in the 30 days before CTR in ASA physical status (PS) I-II patients. We also examined the patient, procedure, and facility factors associated with receiving ≥1 LVT with mixed-effects logistic regression and the number of tests received with mixed-effects negative binomial regression.ResultsFrom FY15-17, 10,000 ASA class I-II patients received a CTR by 699 surgeons in 125 VHA facilities. Overall, 47.0% of patients had a CTR that was preceded by ≥1 LVT, with substantial variability between facilities (range = 0%-100%; interquartile range = 36.3%), representing $339,717 in costs. Older age and female sex were associated with higher odds of receiving ≥1 LVT. Local versus other modes of anesthesia were associated with lower odds of receiving ≥1 LVT. Several facilities experienced large (>25%) increases or decreases from FY15 to FY17 in the proportion of patients receiving ≥1 LVT.ConclusionsCounter to guidance from the ASA, we found that almost half of CTRs performed on ASA class I-II VHA patients were preceded by ≥1 LVT. Although the total cost of these tests is relatively modest, CTR is just one of many low-risk procedures (eg, trigger finger release, cataract surgery) that may involve similar preoperative testing practices. These results will inform site selection for qualitative investigation of the drivers of low-value testing and the development of interventions to improve preoperative testing practice, especially in locations where rates of LVT are high.

Project description:Carpal tunnel syndrome is the most common entrapment neuropathy and is associated with altered brain function and structure. However, little is understood of the central mechanisms associated with its pain, symptom presentation, and treatment-related resolution. This longitudinal study evaluated carpal tunnel syndrome-related alterations in brain network communication and relationships to behavioural signs of central sensitization before and after carpal tunnel release surgery. We tested the hypothesis that carpal tunnel syndrome is associated with condition- and treatment-related plasticity in brain regions involved in somatosensation. We used quantitative sensory testing and clinical and pain questionnaires to assess sensory and pain function in 25 patients with carpal tunnel syndrome before (18 women, 7 men) and after (n = 16) surgery, and 25 sex- and age-matched healthy controls. We also acquired resting-state functional MRI to determine functional connectivity of two key nodes in the somatosensory system, the thalamus and primary somatosensory cortex. Seed-to-whole brain resting-state static functional connectivity analyses revealed abnormally low functional connectivity for the hand area of the primary somatosensory cortex with the contralateral somatosensory association cortex (supramarginal gyrus) before surgery (P < 0.01). After clinically effective surgery: (i) Primary somatosensory functional connectivity was normalized with the contralateral somatosensory association cortex and reduced with the dorsolateral prefrontal cortex (a region associated with cognitive and emotional modulation of pain) and primary visual areas (P < 0.001) from pre-op levels; and (ii) Functional connectivity of the thalamus with the primary somatosensory and motor cortices was attenuated from pre-op levels (P < 0.001) but did not correlate with temporal summation of pain (a behavioural measure of central sensitization) or clinical measures. This study is the first to reveal treatment-related neuroplasticity in resting-state functional connectivity of the somatosensory system in carpal tunnel syndrome. The findings of dysfunctional resting-state functional connectivity point to aberrant neural synchrony between the brain's representation of the hand with regions involved in processing and integrating tactile and nociceptive stimuli and proprioception in carpal tunnel syndrome. Aberrant neural communication between the primary somatosensory hand area and the dorsolateral prefrontal cortex could reflect increased attention to pain, paraesthesia, and altered sensation in the hand. Finally, reduced thalamocortical functional connectivity after surgery may reflect central plasticity in response to the resolution of abnormal sensory signals from the periphery. Our findings support the concept of underlying brain contributions to this peripheral neuropathy, specifically aberrant thalamocortical and corticocortical communication, and point to potential central therapeutic targets to complement peripheral treatments.

Dataset Information

Developing Machine Learning Algorithms to Support Patient-centered, Value-based Carpal Tunnel Decompression Surgery.

Background

Methods

Results

Conclusions

Publications

Developing Machine Learning Algorithms to Support Patient-centered, Value-based Carpal Tunnel Decompression Surgery.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets