Publications by authors named "Tina Hernandez-Boussard"

172 Publications

Disparity in the Setting of Incident Heart Failure Diagnosis.

Circ Heart Fail 2021 Jul 27:CIRCHEARTFAILURE121008538. Epub 2021 Jul 27.

Division of Cardiovascular Medicine and the Cardiovascular Institute, Department of Medicine (A.T.S., F.R., D.J.M., E.L., P.A.H.), Stanford University, CA.

Background: Early heart failure (HF) recognition can reduce morbidity, yet HF is often initially diagnosed only after a patient clinically worsens. We sought to identify characteristics that predict diagnosis in the acute care setting versus the outpatient setting.

Methods: We estimated the proportion of incident HF diagnosed in the acute care setting (inpatient hospital or emergency department) versus outpatient setting based on diagnostic codes from a claims database covering commercial insurance and Medicare Advantage between 2003 and 2019. After excluding new-onset HF potentially caused by a concurrent acute cause (eg, acute myocardial infarction), we identified demographic, clinical, and socioeconomic predictors of diagnosis setting. Patients were linked to their primary care clinicians to evaluate diagnosis setting variation across clinicians.

Results: Of 959 438 patients with new HF, 38% were diagnosed in acute care. Of these, 46% had potential HF symptoms in the prior 6 months. Over time, the relative odds of acute care diagnosis increased by 3.2% annually after adjustment for patient characteristics (95% CI, 3.1%-3.3%). Acute care diagnosis setting was more likely for women compared with men (adjusted odds ratio, 1.11 [95% CI, 1.10-1.12]) and for Black patients compared with White patients (adjusted odds ratio, 1.18 [95% CI, 1.16-1.19]). The proportion of acute care diagnosis varied substantially (interquartile range: 24%-39%) among clinicians after adjusting for patient-level risk factors.

Conclusions: A large proportion of first HF diagnoses occur in the acute care setting, particularly among women and Black patients, yet many had potential HF symptoms in the months before acute care visits. These results raise concerns that many HF diagnoses are missed in the outpatient setting. Earlier diagnosis could allow for timelier high-value interventions, addressing disparities and reducing the progression of HF.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1161/CIRCHEARTFAILURE.121.008538DOI Listing
July 2021

Diverse patient trajectories during cytotoxic chemotherapy: Capturing longitudinal patient-reported outcomes.

Cancer Med 2021 Jul 13. Epub 2021 Jul 13.

Department of Medicine (Biomedical Informatics, Stanford University School of Medicine, Stanford, California, USA.

Background: High-value cancer care balances effective treatment with preservation of quality of life. Chemotherapy is known to affect patients' physical and psychological well-being negatively. Patient-reported outcomes (PROs) provide a means to monitor declines in a patients' well-being during treatment.

Methods: We identified 741 oncology patients undergoing chemotherapy in our electronic health record (EHR) system who completed Patient-Reported Outcomes Measurement Information System (PROMIS) surveys during treatment at a comprehensive cancer center, 2013-2018. PROMIS surveys were collected before, during, and after chemotherapy treatment. Linear mixed-effects models were performed to identify predictors of physical and mental health scores over time. A k-mean cluster analysis was used to group patient PROMIS score trajectories.

Results: Mean global physical health (GPH) scores were 48.7 (SD 9.3), 47.7 (8.8), and 48.6 (8.9) and global mental health (GMH) scores were 50.4 (8.6), 49.5 (8.8), and 50.6 (9.1) before, during, and after chemotherapy, respectively. Asian race, Hispanic ethnicity, public insurance, anxiety/depression, stage III cancer, and palliative care were predictors of GPH and GMH decline. The treatment time period was also a predictor of both GPH and GMH decline relative to pre-treatment. Trajectory clustering identified four distinct PRO clusters associated with chemotherapy treatment.

Conclusions: Patient-reported outcomes are increasingly used to help monitor cancer treatment and are now a part of care reimbursement. This study leveraged routinely collected PROMIS surveys linked to EHRs to identify novel patient trajectories of physical and mental well-being in oncology patients undergoing chemotherapy and potential predictors. Supportive care interventions in high-risk populations identified by our study may optimize resource deployment.

Novelty And Impact: This study leveraged routinely collected patient-reported outcome (PROMIS) surveys linked to electronic health records to characterize oncology patients' quality of life during chemotherapy. Important clinical and demographic predictors of declines in quality of life were identified and four novel trajectories to guide personalized interventions and support. This work highlights the utility of monitoring patient-reported outcomes not only before and after, but during chemotherapy to help advert adverse patient outcomes and improve treatment adherence.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/cam4.4124DOI Listing
July 2021

Increases in SARS-CoV-2 Test Positivity Rates Among Hispanic People in a Northern California Health System.

Public Health Rep 2021 Jun 23:333549211026778. Epub 2021 Jun 23.

Department of Medicine, Stanford University, Stanford, CA, USA.

Racial/ethnic minority groups are disproportionately affected by the COVID-19 pandemic. We examined ethnic differences in SARS-CoV-2 testing patterns and positivity rates in a large health care system in Northern California. The study population included patients tested for SARS-CoV-2 from March 4, 2020, through January 12, 2021, at Stanford Health Care. We used adjusted hierarchical logistic regression models to identify factors associated with receiving a positive test result. During the study period, 282 916 SARS-CoV-2 tests were administered to 179 032 unique patients, 32 766 (18.3%) of whom were Hispanic. Hispanic patients were 3 times more likely to receive a positive test result than patients in other racial/ethnic groups (odds ratio = 3.16; 95% CI, 3.00-3.32). The rate of receiving a positive test result for SARS-CoV-2 among Hispanic patients increased from 5.4% in mid-March to 15.7% in mid-July, decreased to 3.9% in mid-October, and increased to 21.2% toward the end of December. Hispanic patients were more likely than non-Hispanic patients to receive a positive test result for SARS-CoV-2, with increasing trends during regional surges. The disproportionate and growing overrepresentation of Hispanic people receiving a positive test result for SARS-CoV-2 demonstrates the need to focus public health prevention efforts on these communities.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/00333549211026778DOI Listing
June 2021

Evaluation of clustering and topic modeling methods over health-related tweets and emails.

Artif Intell Med 2021 07 7;117:102096. Epub 2021 May 7.

Department of Health Outcomes and Biomedical Informatics, College of Medicine, University of Florida, Gainesville, FL, USA.

Background: Internet provides different tools for communicating with patients, such as social media (e.g., Twitter) and email platforms. These platforms provided new data sources to shed lights on patient experiences with health care and improve our understanding of patient-provider communication. Several existing topic modeling and document clustering methods have been adapted to analyze these new free-text data automatically. However, both tweets and emails are often composed of short texts; and existing topic modeling and clustering approaches have suboptimal performance on these short texts. Moreover, research over health-related short texts using these methods has become difficult to reproduce and benchmark, partially due to the absence of a detailed comparison of state-of-the-art topic modeling and clustering methods on these short texts.

Methods: We trained eight state-of- the-art topic modeling and clustering algorithms on short texts from two health-related datasets (tweets and emails): Latent Semantic Indexing (LSI), Latent Dirichlet Allocation (LDA), LDA with Gibbs Sampling (GibbsLDA), Online LDA, Biterm Model (BTM), Online Twitter LDA, and Gibbs Sampling for Dirichlet Multinomial Mixture (GSDMM), as well as the k-means clustering algorithm with two different feature representations: TF-IDF and Doc2Vec. We used cluster validity indices to evaluate the performance of topic modeling and clustering: two internal indices (i.e. assessing the goodness of a clustering structure without external information) and five external indices (i.e. comparing the results of a cluster analysis to an externally known provided class labels).

Results: In overall, for number of clusters (k) from 2 to 50, Online Twitter LDA and GSDMM achieved the best performance in terms of internal indices, while LSI and k-means with TF-IDF had the highest external indices. Also, of all tweets (N = 286, 971; HPV represents 94.6% of tweets and lynch syndrome represents 5.4%), for k = 2, most of the methods could respect this initial clustering distribution. However, we found model performance varies with the source of data and hyper-parameters such as the number of topics and the number of iterations used to train the models. We also conducted an error analysis using the Hamming loss metric, for which the poorest value was obtained by GSDMM on both datasets.

Conclusions: Researchers hoping to group or classify health related short-text data can expect to select the most suitable topic modeling and clustering methods for their specific research questions. Therefore, we presented a comparison of the most common used topic modeling and clustering algorithms over two health-related, short-text datasets using both internal and external clustering validation indices. Internal indices suggested Online Twitter LDA and GSDMM as the best, while external indices suggested LSI and k-means with TF-IDF as the best. In summary, our work suggested researchers can improve their analysis of model performance by using a variety of metrics, since there is not a single best metric.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.artmed.2021.102096DOI Listing
July 2021

Learning from past respiratory failure patients to triage COVID-19 patient ventilator needs: A multi-institutional study.

J Biomed Inform 2021 07 27;119:103802. Epub 2021 May 27.

Department of Medicine, Stanford University, Stanford, CA, United States; Department of Biomedical Data Science, Stanford University, Stanford, CA, United States; Department of Surgery, Stanford University, Stanford, CA, United States. Electronic address:

Background: Unlike well-established diseases that base clinical care on randomized trials, past experiences, and training, prognosis in COVID19 relies on a weaker foundation. Knowledge from other respiratory failure diseases may inform clinical decisions in this novel disease. The objective was to predict 48-hour invasive mechanical ventilation (IMV) within 48 h in patients hospitalized with COVID-19 using COVID-like diseases (CLD).

Methods: This retrospective multicenter study trained machine learning (ML) models on patients hospitalized with CLD to predict IMV within 48 h in COVID-19 patients. CLD patients were identified using diagnosis codes for bacterial pneumonia, viral pneumonia, influenza, unspecified pneumonia and acute respiratory distress syndrome (ARDS), 2008-2019. A total of 16 cohorts were constructed, including any combinations of the four diseases plus an exploratory ARDS cohort, to determine the most appropriate cohort to use. Candidate predictors included demographic and clinical parameters that were previously associated with poor COVID-19 outcomes. Model development included the implementation of logistic regression and three ensemble tree-based algorithms: decision tree, AdaBoost, and XGBoost. Models were validated in hospitalized COVID-19 patients at two healthcare systems, March 2020-July 2020. ML models were trained on CLD patients at Stanford Hospital Alliance (SHA). Models were validated on hospitalized COVID-19 patients at both SHA and Intermountain Healthcare.

Results: CLD training data were obtained from SHA (n = 14,030), and validation data included 444 adult COVID-19 hospitalized patients from SHA (n = 185) and Intermountain (n = 259). XGBoost was the top-performing ML model, and among the 16 CLD training cohorts, the best model achieved an area under curve (AUC) of 0.883 in the validation set. In COVID-19 patients, the prediction models exhibited moderate discrimination performance, with the best models achieving an AUC of 0.77 at SHA and 0.65 at Intermountain. The model trained on all pneumonia and influenza cohorts had the best overall performance (SHA: positive predictive value (PPV) 0.29, negative predictive value (NPV) 0.97, positive likelihood ratio (PLR) 10.7; Intermountain: PPV, 0.23, NPV 0.97, PLR 10.3). We identified important factors associated with IMV that are not traditionally considered for respiratory diseases.

Conclusions: The performance of prediction models derived from CLD for 48-hour IMV in patients hospitalized with COVID-19 demonstrate high specificity and can be used as a triage tool at point of care. Novel predictors of IMV identified in COVID-19 are often overlooked in clinical practice. Lessons learned from our approach may assist other research institutes seeking to build artificial intelligence technologies for novel or rare diseases with limited data for training and validation.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jbi.2021.103802DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8159260PMC
July 2021

Development and evaluation of novel ophthalmology domain-specific neural word embeddings to predict visual prognosis.

Int J Med Inform 2021 06 16;150:104464. Epub 2021 Apr 16.

Center for Biomedical Informatics Research, School of Medicine, Stanford University, 1265 Welch Road, Stanford, CA, 94305, United States. Electronic address:

Objective: To develop and evaluate novel word embeddings (WEs) specific to ophthalmology, using text corpora from published literature and electronic health records (EHR).

Materials And Methods: We trained ophthalmology-specific WEs using 121,740 PubMed abstracts and 89,282 EHR notes using word2vec continuous bag-of-words architecture. PubMed and EHR WEs were compared to general domain GloVe WEs and general biomedical domain BioWordVec embeddings using a novel ophthalmology-domain-specific 200-question analogy test and prediction of prognosis in 5547 low vision patients using EHR notes as inputs to a deep learning model.

Results: We found that many words representing important ophthalmic concepts in the EHR were missing from the general domain GloVe vocabulary, but covered in the ophthalmology abstract corpus. On ophthalmology analogy testing, PubMed WEs scored 95.0 %, outperforming EHR (86.0 %) and GloVe (91.0 %) but less than BioWordVec (99.5 %). On predicting low vision prognosis, PubMed and EHR WEs resulted in similar AUROC (0.830; 0.826), outperforming GloVe (0.778) and BioWordVec (0.784).

Conclusion: We found that using ophthalmology domain-specific WEs improved performance in ophthalmology-related clinical prediction compared to general WEs. Deep learning models using clinical notes as inputs can predict the prognosis of visually impaired patients. This work provides a framework to improve predictive models using domain-specific WEs.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.ijmedinf.2021.104464DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8183292PMC
June 2021

Prevalence of Postprostatectomy Incontinence Requiring Anti-incontinence Surgery After Radical Prostatectomy for Prostate Cancer: A Retrospective Population-Based Analysis.

Int Neurourol J 2021 Mar 9. Epub 2021 Mar 9.

Department of Urology, Stanford University Medical Center, Stanford, CA, USA.

Purpose: The aim of this study was to examine the prevalence of surgery for post-prostatectomy incontinence (PI) following minimally invasive surgery compared to conventional open surgery for prostate cancer.

Methods: This retrospective cohort study used the Florida State Ambulatory Surgery and State Inpatient Databases, 2008 to 2010, RP patients were identified using ICD-9/10 procedure codes and among this cohort PI was identified also using ICD-9/10 codes. Surgical approaches included Minimally invasive (robotic or laparoscopic) vs. open (retropubic or perineal) RP. The primary outcome was the overall prevalence of surgery for PI. The secondary outcome was the association of PI requiring anti-incontinence surgery with the surgical approach for RP.

Results: Among the 13535 patients initially included in the study (mean age, 63.3 years), 6932 (51.2%) underwent open RP and 6603 (49.8%) underwent minimally invasive RP. The overall prevalence of surgical procedures for PI during the observation period among the all patients who had received RP was 3.3%. The rate of PI surgery for patients receiving minimally invasive surgery was higher than that for patients receiving open surgery (4.8% vs. 3.0%; risk difference, 1.8%; 95% CI, 0.3% to 3.4%). The adjusted prevalence of PI surgery for patients who had undergone laparoscopic RP was higher than that for those with retropubic RP (8.6% vs. 3.7%).

Conclusions: Among patients undergoing RP for prostate cancer, the prevalence of PI surgery is not negligible. Patients undergoing minimally invasive RP had higher adjusted rates for PI surgery compared to open approaches, which was attributed to high rate of PI surgery following laparoscopic approach and low rate of PI surgery following perineal approach. More studies are needed to establish strategies to reduce the rate of PI surgery after RP.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.5213/inj.2040296.148DOI Listing
March 2021

Conflicting information from the Food and Drug Administration: Missed opportunity to lead standards for safe and effective medical artificial intelligence solutions.

J Am Med Inform Assoc 2021 06;28(6):1353-1355

Department of Medicine, Stanford University, Stanford, California, USA.

The Food & Drug Administration (FDA) is considering the permanent exemption of premarket notification requirements for several Class I and II medical device products, including several artificial Intelligence (AI)-driven devices. The exemption is based on the need to rapidly more quickly disseminate devices to the public, estimated cost-savings, a lack of documented adverse events reported to the FDA's database. However, this ignores emerging issues related to AI-based devices, including utility, reproducibility and bias that may not only affect an individual but entire populations. We urge the FDA to reinforce the messaging on safety and effectiveness regulations of AI-based Software as a Medical Device products to better promote fair AI-driven clinical decision tools and for preventing harm to the patients we serve.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1093/jamia/ocab035DOI Listing
June 2021

Learning From Past Respiratory Infections to Predict COVID-19 Outcomes: Retrospective Study.

J Med Internet Res 2021 02 22;23(2):e23026. Epub 2021 Feb 22.

Department of Medicine, Biomedical Informatics, Stanford University, Stanford, CA, United States.

Background: For the clinical care of patients with well-established diseases, randomized trials, literature, and research are supplemented with clinical judgment to understand disease prognosis and inform treatment choices. In the void created by a lack of clinical experience with COVID-19, artificial intelligence (AI) may be an important tool to bolster clinical judgment and decision making. However, a lack of clinical data restricts the design and development of such AI tools, particularly in preparation for an impending crisis or pandemic.

Objective: This study aimed to develop and test the feasibility of a "patients-like-me" framework to predict the deterioration of patients with COVID-19 using a retrospective cohort of patients with similar respiratory diseases.

Methods: Our framework used COVID-19-like cohorts to design and train AI models that were then validated on the COVID-19 population. The COVID-19-like cohorts included patients diagnosed with bacterial pneumonia, viral pneumonia, unspecified pneumonia, influenza, and acute respiratory distress syndrome (ARDS) at an academic medical center from 2008 to 2019. In total, 15 training cohorts were created using different combinations of the COVID-19-like cohorts with the ARDS cohort for exploratory purposes. In this study, two machine learning models were developed: one to predict invasive mechanical ventilation (IMV) within 48 hours for each hospitalized day, and one to predict all-cause mortality at the time of admission. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, positive predictive value, and negative predictive value. We established model interpretability by calculating SHapley Additive exPlanations (SHAP) scores to identify important features.

Results: Compared to the COVID-19-like cohorts (n=16,509), the patients hospitalized with COVID-19 (n=159) were significantly younger, with a higher proportion of patients of Hispanic ethnicity, a lower proportion of patients with smoking history, and fewer patients with comorbidities (P<.001). Patients with COVID-19 had a lower IMV rate (15.1 versus 23.2, P=.02) and shorter time to IMV (2.9 versus 4.1 days, P<.001) compared to the COVID-19-like patients. In the COVID-19-like training data, the top models achieved excellent performance (AUROC>0.90). Validating in the COVID-19 cohort, the top-performing model for predicting IMV was the XGBoost model (AUROC=0.826) trained on the viral pneumonia cohort. Similarly, the XGBoost model trained on all 4 COVID-19-like cohorts without ARDS achieved the best performance (AUROC=0.928) in predicting mortality. Important predictors included demographic information (age), vital signs (oxygen saturation), and laboratory values (white blood cell count, cardiac troponin, albumin, etc). Our models had class imbalance, which resulted in high negative predictive values and low positive predictive values.

Conclusions: We provided a feasible framework for modeling patient deterioration using existing data and AI technology to address data limitations during the onset of a novel, rapidly changing pandemic.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2196/23026DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7901593PMC
February 2021

Assessment of a Clinical Trial-Derived Survival Model in Patients With Metastatic Castration-Resistant Prostate Cancer.

JAMA Netw Open 2021 01 4;4(1):e2031730. Epub 2021 Jan 4.

Department of Medicine, Stanford University School of Medicine, Stanford, California.

Importance: Randomized clinical trials (RCTs) are considered the criterion standard for clinical evidence. Despite their many benefits, RCTs have limitations, such as costliness, that may reduce the generalizability of their findings among diverse populations and routine care settings.

Objective: To assess the performance of an RCT-derived prognostic model that predicts survival among patients with metastatic castration-resistant prostate cancer (CRPC) when the model is applied to real-world data from electronic health records (EHRs).

Design, Setting, And Participants: The RCT-trained model and patient data from the RCTs were obtained from the Dialogue for Reverse Engineering Assessments and Methods (DREAM) challenge for prostate cancer, which occurred from March 16 to July 27, 2015. This challenge included 4 phase 3 clinical trials of patients with metastatic CRPC. Real-world data were obtained from the EHRs of a tertiary care academic medical center that includes a comprehensive cancer center. In this study, the DREAM challenge RCT-trained model was applied to real-world data from January 1, 2008, to December 31, 2019; the model was then retrained using EHR data with optimized feature selection. Patients with metastatic CRPC were divided into RCT and EHR cohorts based on data source. Data were analyzed from March 23, 2018, to October 22, 2020.

Exposures: Patients who received treatment for metastatic CRPC.

Main Outcomes And Measures: The primary outcome was the performance of an RCT-derived prognostic model that predicts survival among patients with metastatic CRPC when the model is applied to real-world data. Model performance was compared using 10-fold cross-validation according to time-dependent integrated area under the curve (iAUC) statistics.

Results: Among 2113 participants with metastatic CRPC, 1600 participants were included in the RCT cohort, and 513 participants were included in the EHR cohort. The RCT cohort comprised a larger proportion of White participants (1390 patients [86.9%] vs 337 patients [65.7%]) and a smaller proportion of Hispanic participants (14 patients [0.9%] vs 42 patients [8.2%]), Asian participants (41 patients [2.6%] vs 88 patients [17.2%]), and participants older than 75 years (388 patients [24.3%] vs 191 patients [37.2%]) compared with the EHR cohort. Participants in the RCT cohort also had fewer comorbidities (mean [SD], 1.6 [1.8] comorbidities vs 2.5 [2.6] comorbidities, respectively) compared with those in the EHR cohort. Of the 101 variables used in the RCT-derived model, 10 were not available in the EHR data set, 3 of which were among the top 10 features in the DREAM challenge RCT model. The best-performing EHR-trained model included only 25 of the 101 variables included in the RCT-trained model. The performance of the RCT-trained and EHR-trained models was adequate in the EHR cohort (mean [SD] iAUC, 0.722 [0.118] and 0.762 [0.106], respectively); model optimization was associated with improved performance of the best-performing EHR model (mean [SD] iAUC, 0.792 [0.097]). The EHR-trained model classified 256 patients as having a high risk of mortality and 256 patients as having a low risk of mortality (hazard ratio, 2.7; 95% CI, 2.0-3.7; log-rank P < .001).

Conclusions And Relevance: In this study, although the RCT-trained models did not perform well when applied to real-world EHR data, retraining the models using real-world EHR data and optimizing variable selection was beneficial for model performance. As clinical evidence evolves to include more real-world data, both industry and academia will likely search for ways to balance model optimization with generalizability. This study provides a pragmatic approach to applying RCT-trained models to real-world data.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1001/jamanetworkopen.2020.31730DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7823224PMC
January 2021

Real-world Evidence to Estimate Prostate Cancer Costs for First-line Treatment or Active Surveillance.

Eur Urol Open Sci 2021 Jan 10;23:20-29. Epub 2020 Dec 10.

Department of Medicine, School of Medicine, Stanford University, Stanford, CA, USA.

Background: Prostate cancer is the most common cancer in men and second leading cause of cancer-related deaths. Changes in screening guidelines, adoption of active surveillance (AS), and implementation of high-cost technologies have changed treatment costs. Traditional cost-effectiveness studies rely on clinical trial protocols unlikely to capture actual practice behavior, and existing studies use data predating new technologies. Real-world evidence reflecting these changes is lacking.

Objective: To assess real-world costs of first-line prostate cancer management.

Design Setting And Participants: We used clinical electronic health records for 2008-2018 linked with the California Cancer Registry and the Medicare Fee Schedule to assess costs over 24 or 60 mo following diagnosis. We identified surgery or radiation treatments with structured methods, while we used both structured data and natural language processing to identify AS.

Outcome Measurements And Statistical Analysis: Our results are risk-stratified calculated cost per day (CCPD) for first-line management, which are independent of treatment duration. We used the Kruskal-Wallis test to compare unadjusted CCPD while analysis of covariance log-linear models adjusted estimates for age and Charlson comorbidity.

Results And Limitations: In 3433 patients, surgery (54.6%) was more common than radiation (22.3%) or AS (23.0%). Two years following diagnosis, AS ($2.97/d) was cheaper than surgery ($5.67/d) or radiation ($9.34/d) in favorable disease, while surgery ($7.17/d) was cheaper than radiation ($16.34/d) for unfavorable disease. At 5 yr, AS ($2.71/d) remained slightly cheaper than surgery ($2.87/d) and radiation ($4.36/d) in favorable disease, while for unfavorable disease surgery ($4.15/d) remained cheaper than radiation ($10.32/d). Study limitations include information derived from a single healthcare system and costs based on benchmark Medicare estimates rather than actual payment exchanges.

Patient Summary: Active surveillance was cheaper than surgery (-47.6%) and radiation (-68.2%) at 2 yr for favorable-risk disease, which decreased by 5 yr (-5.6% and -37.8%, respectively). Surgery was less costly than radiation for unfavorable risk for both intervals (-56.1% and -59.8%, respectively).
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.euros.2020.11.004DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7751921PMC
January 2021

Contemporary Practices and Complications of Surgery for Thoracic Outlet Syndrome in the United States.

Ann Vasc Surg 2021 Apr 3;72:147-158. Epub 2021 Feb 3.

Division of Vascular & Endovascular Surgery, Department of Surgery, Stanford University School of Medicine, Stanford, CA. Electronic address:

Background: Thoracic outlet syndrome (TOS) surgery is relatively rare and controversial, given the challenges in diagnosis as well as wide variation in symptomatic and functional recovery. Our aims were to measure trends in utilization of TOS surgery, complications, and mortality rates in a nationally representative cohort and compare higher versus lower volume centers.

Methods: The National Inpatient Sample was queried using International Classification of Diseases, Ninth Revision, codes for rib resection and scalenectomy paired with axillo-subclavian aneurysm (arterial [aTOS]), subclavian deep vein thrombosis (venous [vTOS]), or brachial plexus lesions (neurogenic [nTOS]). Basic descriptive statistics, nonparametric tests for trend, and multivariable hierarchical regression models with random intercept for center were used to compare outcomes for TOS types, trends over time, and higher and lower volume hospitals, respectively.

Results: There were 3,547 TOS operations (for an estimated 18,210 TOS operations nationally) performed between 2010 and 2015 (89.2% nTOS, 9.9% vTOS, and 0.9% aTOS) with annual case volume increasing significantly over time (P = 0.03). Higher volume centers (≥10 cases per year) represented 5.2% of hospitals and 37.0% of cases, and these centers achieved significantly lower overall major complication (defined as neurologic injury, arterial or venous injury, vascular graft complication, pneumothorax, hemorrhage/hematoma, or lymphatic leak) rates (adjusted odds ratio [OR] 0.71 [95% confidence interval 0.52-0.98]; P = 0.04], but no difference in neurologic complications such as brachial plexus injury (aOR 0.69 [0.20-2.43]; P = 0.56) or vascular injuries/graft complications (aOR 0.71 [0.0.33-1.54]; P = 0.39). Overall mortality was 0.6%, neurologic injury was rare (0.3%), and the proportion of patients experiencing complications decreased over time (P = 0.03). However, vTOS and aTOS had >2.5 times the odds of major complication compared with nTOS (OR 2.68 [1.88-3.82] and aOR 4.26 [1.78-10.17]; P < 0.001), and ∼10 times the odds of a vascular complication (aOR 10.37 [5.33-20.19] and aOR 12.93 [3.54-47.37]; P < 0.001], respectively. As the number of complications decreased, average hospital charges also significantly decreased over time (P < 0.001). Total hospital charges were on average higher when surgery was performed in lower volume centers (<10 cases per year) compared with higher volume centers (mean $65,634 [standard deviation 98,796] vs. $45,850 [59,285]; P < 0.001).

Conclusions: The annual number of TOS operations has increased in the United States from 2010 to 2015, whereas complications and average hospital charges have decreased. Mortality and neurologic injury remain rare. Higher volume centers delivered higher value care: less or similar operative morbidity with lower total hospital charges.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.avsg.2020.10.046DOI Listing
April 2021

Preoperative Factors Associated with Remote Postoperative Pain Resolution and Opioid Cessation in a Mixed Surgical Cohort: Post Hoc Analysis of a Perioperative Gabapentin Trial.

J Pain Res 2020 18;13:2959-2970. Epub 2020 Nov 18.

Division of Pain Medicine, Department of Anesthesiology, Perioperative, and Pain Medicine, Stanford University, Stanford, CA, USA.

Background: Preoperative patient-specific risk factors may elucidate the mechanisms leading to the persistence of pain and opioid use after surgery. This study aimed to determine whether similar or discordant preoperative factors were associated with the duration of postoperative pain and opioid use.

Methods: In this post hoc analysis of a randomized, double-blind, placebo-controlled trial of perioperative gabapentin vs active placebo, 410 patients aged 18-75 years, undergoing diverse operations underwent preoperative assessments of pain, opioid use, substance use, and psychosocial variables. After surgery, a modified Brief Pain Inventory was administered over the phone daily up to 3 months, weekly up to 6 months, and monthly up to 2 years after surgery. Pain and opioid cessation were defined as the first of 5 consecutive days of 0 out of 10 pain or no opioid use, respectively.

Results: Overall, 36.1%, 19.8%, and 9.5% of patients continued to report pain, and 9.5%, 2.4%, and 1.7% reported continued opioid use at 3, 6, and 12 months after surgery. Preoperative pain at the future surgical site (every 1-point increase in the Numeric Pain Rating Scale; HR 0.93; 95% CI 0.87-1.00; =0.034), trait anxiety (every 10-point increase in the Trait Anxiety Inventory; HR 0.79; 95% CI 0.68-0.92; =0.002), and a history of delayed recovery after injury (HR 0.62; 95% CI 0.40-0.96; =0.034) were associated with delayed pain cessation. Preoperative opioid use (HR 0.60; 95% CI 0.39-0.92; =0.020), elevated depressive symptoms (every 5-point increase in the Beck Depression Inventory-II score; HR 0.88; 95% CI 0.80-0.98; =0.017), and preoperative pain outside of the surgical site (HR 0.94; 95% CI 0.89-1.00; =0.046) were associated with delayed opioid cessation, while perioperative gabapentin promoted opioid cessation (HR 1.37; 95% CI 1.06-1.77; =0.016).

Conclusion: Separate risk factors for prolonged post-surgical pain and opioid use indicate that preoperative risk stratification for each outcome may identify patients needing personalized care to augment universal protocols for perioperative pain management and conservative opioid prescribing to improve long-term outcomes.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2147/JPR.S269370DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7680674PMC
November 2020

Patient Electronic Health Records Score for Preoperative Risk Assessment Before Total Knee Arthroplasty.

JB JS Open Access 2020 Apr-Jun;5(2):e0061. Epub 2020 May 6.

Palo Alto Veterans Hospital, Palo Alto, California.

Background: Current preoperative risk assessment tools are often cumbersome, have limited accuracy, and are poorly adopted. The Care Assessment Need (CAN) score, an existing tool developed for primary care providers in the U.S. Veterans Administration health-care system (VA), is automatically calculated for individual patients using electronic health record data. Therefore, it could present an efficient preoperative risk assessment tool. The aim of this project was to determine if the CAN score can be repurposed as a preoperative risk assessment tool for patients undergoing total knee arthroplasty (TKA).

Methods: A multicenter retrospective observational study was conducted using national VA data from 2013 to 2016. The cohort included veterans who underwent TKA identified through ICD-9 (International Classification of Diseases, Ninth Revision), ICD-10, and CPT (Current Procedural Terminology) codes. The focus of the study was the preoperative patient CAN score, a single numerical value ranging from 0 to 99 (with a higher score representing greater risk) that is automatically calculated each week using multiple data points in the VA electronic health record. Study outcomes of interest were 90-day readmission, prolonged hospital stay (>5 days), 1-year mortality, and non-routine patient discharge.

Results: The study included 17,210 veterans. Their median preoperative CAN score was 75, although there was substantial variability in patient CAN scores among different facilities. A preoperative CAN score of >75 was significantly associated with mortality (odds ratio [OR] = 3.54), prolonged length of stay (OR = 1.97), 90-day readmission (OR = 1.65), and non-routine discharge (OR = 1.57). The CAN score had good accuracy with a receiver operating characteristic (ROC) curve value of >0.7 for all outcomes except 90-day readmission.

Conclusions: The CAN score can be leveraged as an extremely efficient way to risk-stratify patients before TKA, with results that surpass other commonly available and labor-intensive alternatives. As a result, this simple and efficient solution is well positioned for broad adoption as a standardized decision support tool.

Level Of Evidence: Prognostic Level IV. See Instructions for Authors for a complete description of levels of evidence.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2106/JBJS.OA.19.00061DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7418912PMC
May 2020

Phenotyping severity of patient-centered outcomes using clinical notes: A prostate cancer use case.

Learn Health Syst 2020 Oct 17;4(4):e10237. Epub 2020 Jul 17.

Department of Medicine, Biomedical Informatics Research Stanford University Stanford California USA.

Introduction: A learning health system (LHS) must improve care in ways that are meaningful to patients, integrating patient-centered outcomes (PCOs) into core infrastructure. PCOs are common following cancer treatment, such as urinary incontinence (UI) following prostatectomy. However, PCOs are not systematically recorded because they can only be described by the patient, are subjective and captured as unstructured text in the electronic health record (EHR). Therefore, PCOs pose significant challenges for phenotyping patients. Here, we present a natural language processing (NLP) approach for phenotyping patients with UI to classify their disease into severity subtypes, which can increase opportunities to provide precision-based therapy and promote a value-based delivery system.

Methods: Patients undergoing prostate cancer treatment from 2008 to 2018 were identified at an academic medical center. Using a hybrid NLP pipeline that combines rule-based and deep learning methodologies, we classified positive UI cases as mild, moderate, and severe by mining clinical notes.

Results: The rule-based model accurately classified UI into disease severity categories (accuracy: 0.86), which outperformed the deep learning model (accuracy: 0.73). In the deep learning model, the recall rates for mild and moderate group were higher than the precision rate (0.78 and 0.79, respectively). A hybrid model that combined both methods did not improve the accuracy of the rule-based model but did outperform the deep learning model (accuracy: 0.75).

Conclusion: Phenotyping patients based on indication and severity of PCOs is essential to advance a patient centered LHS. EHRs contain valuable information on PCOs and by using NLP methods, it is feasible to accurately and efficiently phenotype PCO severity. Phenotyping must extend beyond the identification of disease to provide classification of disease severity that can be used to guide treatment and inform shared decision-making. Our methods demonstrate a path to a patient centered LHS that could advance precision medicine.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/lrh2.10237DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7556418PMC
October 2020

Clinical Trial Outcomes in Urology: Assessing Early Discontinuation, Results Reporting and Publication in ClinicalTrials.Gov Registrations 2007-2019.

J Urol 2021 04 20;205(4):1159-1168. Epub 2020 Oct 20.

Department of Urology, Stanford University, Stanford, California.

Purpose: Clinical trials require significant resources, but benefits are only realized after trial completion and dissemination of results. We comprehensively assessed early discontinuation, registry results reporting, and publication by trial sponsor and subspecialty in urology trials.

Materials And Methods: We assessed trial registrations from 2007 to 2019 on ClinicalTrials.gov and publication data from PubMed®/MEDLINE®. Associations between sponsor or subspecialty with early discontinuation were assessed using Cox proportional hazards and results reporting or publication with logistic regression at 3 years after completion.

Results: Of 8,636 trials 3,541 (41.0%) were completed and 999 (11.6%) were discontinued. Of completed trials 26.9% reported results and 21.6% were published. Sponsors included academic institutions (53.1%), industry (37.1%) and the U.S. government (9.8%). Academic-sponsored (adjusted HR 0.81, 95% CI 0.69-0.96, p=0.012) and government-sponsored trials (adjusted HR 0.62, 95% CI 0.49-0.78, p <0.001) were less likely than industry to discontinue early. Government-sponsored trials were more likely to report (adjusted OR 1.72, 95% CI 1.17-2.54, p=0.006) and publish (adjusted OR 1.89, 95% CI 1.23-2.89, p=0.004). Academic-sponsored trials were less likely to report (adjusted OR 0.65, CI:0.48-0.88, p=0.006) but more likely to publish (adjusted OR 1.72, 95% CI 1.25-2.37, p <0.001). These outcomes were similar across subspecialties. However, endourology was more likely to discontinue early (adjusted HR 2.00, 95% CI 1.53-2.95, p <0.001), general urology was more likely to report results (adjusted OR 1.54, 95% CI 1.13-2.11, p=0.006) and andrology was less likely to publish (adjusted OR 0.53, 95% CI 0.35-0.81, p=0.003).

Conclusions: Sponsor type is significantly associated with trial completion and dissemination. Government-sponsored trials had the best performance, while industry and academic-sponsored trials lagged in completion and results reporting, respectively. Subspecialty played a lesser role. Lack of dissemination remains a problem for urology trials.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1097/JU.0000000000001432DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8296852PMC
April 2021

Association between patient-initiated emails and overall 2-year survival in cancer patients undergoing chemotherapy: Evidence from the real-world setting.

Cancer Med 2020 11 28;9(22):8552-8561. Epub 2020 Sep 28.

Department of Medicine, Stanford University, Stanford, CA, USA.

Purpose: Prior studies suggest email communication between patients and providers may improve patient engagement and health outcomes. The purpose of this study was to determine whether patient-initiated emails are associated with overall survival benefits among cancer patients undergoing chemotherapy.

Patients And Methods: We identified patient-initiated emails through the patient portal in electronic health records (EHR) among 9900 cancer patients receiving chemotherapy between 2013 and 2018. Email users were defined as patients who sent at least one email 12 months before to 2 months after chemotherapy started. A propensity score-matched cohort analysis was carried out to reduce bias due to confounding (age, primary cancer type, gender, insurance payor, ethnicity, race, stage, income, Charlson score, county of residence). The cohort included 3223 email users and 3223 non-email users. The primary outcome was overall 2-year survival stratified by email use. Secondary outcomes included number of face-to-face visits, prescriptions, and telephone calls. The healthcare teams' response to emails and other forms of communication was also investigated. Finally, a quality measure related to chemotherapy-related inpatient and emergency department visits was evaluated.

Results: Overall 2-year survival was higher in patients who were email users, with an adjusted hazard ratio of 0.80 (95 CI 0.72-0.90; p < 0.001). Email users had higher rates of healthcare utilization, including face-to-face visits (63 vs. 50; p < 0.001), drug prescriptions (28 vs. 21; p < 0.001), and phone calls (18 vs. 16; p < 0.001). Clinical quality outcome measure of inpatient use was better among email users (p = 0.015).

Conclusion: Patient-initiated emails are associated with a survival benefit among cancer patients receiving chemotherapy and may be a proxy for patient engagement. As value-based payment models emphasize incorporating the patients' voice into their care, email communications could serve as a novel source of patient-generated data.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/cam4.3483DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7666724PMC
November 2020

Reporting of demographic data and representativeness in machine learning models using electronic health records.

J Am Med Inform Assoc 2020 12;27(12):1878-1884

Department of Medicine, Stanford University, Stanford, California, USA.

Objective: The development of machine learning (ML) algorithms to address a variety of issues faced in clinical practice has increased rapidly. However, questions have arisen regarding biases in their development that can affect their applicability in specific populations. We sought to evaluate whether studies developing ML models from electronic health record (EHR) data report sufficient demographic data on the study populations to demonstrate representativeness and reproducibility.

Materials And Methods: We searched PubMed for articles applying ML models to improve clinical decision-making using EHR data. We limited our search to papers published between 2015 and 2019.

Results: Across the 164 studies reviewed, demographic variables were inconsistently reported and/or included as model inputs. Race/ethnicity was not reported in 64%; gender and age were not reported in 24% and 21% of studies, respectively. Socioeconomic status of the population was not reported in 92% of studies. Studies that mentioned these variables often did not report if they were included as model inputs. Few models (12%) were validated using external populations. Few studies (17%) open-sourced their code. Populations in the ML studies include higher proportions of White and Black yet fewer Hispanic subjects compared to the general US population.

Discussion: The demographic characteristics of study populations are poorly reported in the ML literature based on EHR data. Demographic representativeness in training data and model transparency is necessary to ensure that ML models are deployed in an equitable and reproducible manner. Wider adoption of reporting guidelines is warranted to improve representativeness and reproducibility.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1093/jamia/ocaa164DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7727384PMC
December 2020

Bias at warp speed: how AI may contribute to the disparities gap in the time of COVID-19.

J Am Med Inform Assoc 2021 01;28(1):190-192

Department of Medicine (Biomedical Informatics), Stanford University, Stanford, California, USA.

The COVID-19 pandemic is presenting a disproportionate impact on minorities in terms of infection rate, hospitalizations, and mortality. Many believe artificial intelligence (AI) is a solution to guide clinical decision-making for this novel disease, resulting in the rapid dissemination of underdeveloped and potentially biased models, which may exacerbate the disparities gap. We believe there is an urgent need to enforce the systematic use of reporting standards and develop regulatory frameworks for a shared COVID-19 data source to address the challenges of bias in AI during this pandemic. There is hope that AI can help guide treatment decisions within this crisis; yet given the pervasiveness of biases, a failure to proactively develop comprehensive mitigation strategies during the COVID-19 pandemic risks exacerbating existing health disparities.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1093/jamia/ocaa210DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7454645PMC
January 2021

Intraocular Pressure Changes after Cataract Surgery in Patients with and without Glaucoma: An Informatics-Based Approach.

Ophthalmol Glaucoma 2020 Sep - Oct;3(5):343-349. Epub 2020 Jun 9.

Byers Eye Institute, Stanford University, Palo Alto, California; Veterans Affairs Palo Alto Health Care System, Palo Alto, California.

Purpose: To evaluate changes in intraocular pressure (IOP) after cataract surgery among patients with or without glaucoma using automated extraction of data from electronic health records (EHRs).

Design: Retrospective cohort study.

Participants: Adults who underwent standalone cataract surgery at a single academic center from 2009-2018.

Methods: Patient information was identified from procedure and billing codes, demographic tables, medication orders, clinical notes, and eye examination fields in the EHR. A previously validated natural language processing pipeline was used to identify laterality of cataract surgery from operative notes and laterality of eye medications from medication orders. Cox proportional hazards modeling evaluated factors associated with the main outcome of sustained postoperative IOP reduction.

Main Outcome Measures: Sustained post-cataract surgery IOP reduction, measured at 14 months or the last follow-up while using equal or fewer glaucoma medications compared with baseline and without additional glaucoma laser or surgery on the operative eye.

Results: The median follow-up for 7574 eyes of 4883 patients who underwent cataract surgery was 244 days. The mean preoperative IOP for all patients was 15.2 mmHg (standard deviation [SD], 3.4 mmHg), which decreased to 14.2 mmHg (SD, 3.0 mmHg) at 12 months after surgery. Patients with IOP of 21.0 mmHg or more showed mean postoperative IOP reduction ranging from -6.2 to -6.9 mmHg. Cataract surgery was more likely to yield sustained IOP reduction for patients with primary open-angle glaucoma (hazard ratio [HR], 1.19; 95% confidence interval, 1.05-1.36) or narrow angles or angle closure (HR, 1.21; 95% confidence interval, 1.08-1.34) compared with patients without glaucoma. Those with a higher baseline IOP were more likely to achieve postoperative IOP reduction (HR, 1.06 per 1-mmHg increase in baseline IOP; 95% confidence interval, 1.05-1.07).

Conclusions: Our results suggest that patients with primary open-angle glaucoma or with narrow angles or chronic angle closure were more likely to achieve sustained IOP reduction after cataract surgery. Patients with higher baseline IOP had increasingly higher odds of achieving reduction in IOP. This evidence demonstrates the potential usefulness of a pipeline for automated extraction of ophthalmic surgical outcomes from EHR to answer key clinical questions on a large scale.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.ogla.2020.06.002DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7529869PMC
June 2020

MINIMAR (MINimum Information for Medical AI Reporting): Developing reporting standards for artificial intelligence in health care.

J Am Med Inform Assoc 2020 12;27(12):2011-2015

Department of Medicine, Stanford University, Stanford, California, USA.

The rise of digital data and computing power have contributed to significant advancements in artificial intelligence (AI), leading to the use of classification and prediction models in health care to enhance clinical decision-making for diagnosis, treatment and prognosis. However, such advances are limited by the lack of reporting standards for the data used to develop those models, the model architecture, and the model evaluation and validation processes. Here, we present MINIMAR (MINimum Information for Medical AI Reporting), a proposal describing the minimum information necessary to understand intended predictions, target populations, and hidden biases, and the ability to generalize these emerging technologies. We call for a standard to accurately and responsibly report on AI in health care. This will facilitate the design and implementation of these models and promote the development and use of associated clinical decision support tools, as well as manage concerns regarding accuracy and bias.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1093/jamia/ocaa088DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7727333PMC
December 2020

Acute pain after breast surgery and reconstruction: A two-institution study of surgical factors influencing short-term pain outcomes.

J Surg Oncol 2020 Jun 20. Epub 2020 Jun 20.

Department of Medicine, Stanford University School of Medicine, Stanford, California.

Background And Objectives: Acute postoperative pain following surgery is known to be associated with chronic pain development and lower quality of life. We sought to analyze the relationship between differing breast cancer excisional procedures, reconstruction, and short-term pain outcomes.

Methods: Women undergoing breast cancer excisional procedures with or without reconstruction at two systems: an academic hospital (AH) and Veterans Health Administration (VHA) were included. Average pain scores at the time of discharge and at 30-day follow-up were analyzed across demographic and clinical characteristics. Linear mixed effects modeling was used to assess the relationship between patient/clinical characteristics and interval pain scores with a random slope to account for differences in baseline pain.

Results: Our study included 1402 patients at AH and 1435 at VHA, of which 426 AH and 165 patients with VHA underwent reconstruction. Pain scores improved over time and were found to be highest at discharge. Time at discharge, 30-day follow-up, and preoperative opioid use were the strongest predictors of high pain scores. Younger age and longer length of stay were independently associated with worse pain scores.

Conclusions: Younger age, preoperative opioid use, and longer length of stay were associated with higher levels of postoperative pain across both sites.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/jso.26070DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7749807PMC
June 2020

Improvement in Patient Safety May Precede Policy Changes: Trends in Patient Safety Indicators in the United States, 2000-2013.

J Patient Saf 2021 Jun;17(4):e327-e334

Department of Biomedical and Neuromotor Sciences, University of Bologna, Bologna, Italy.

Objectives: Quality and safety improvement are global priorities. In the last two decades, the United States has introduced several payment reforms to improve patient safety. The Agency for Healthcare Research and Quality (AHRQ) developed tools to identify preventable inpatient adverse events using administrative data, patient safety indicators (PSIs). The aim of this study was to assess changes in national patient safety trends that corresponded to U.S. pay-for-performance reforms.

Methods: This is a retrospective, longitudinal analysis to estimate temporal changes in 13 AHRQ's PSIs. National inpatient sample from the AHRQ and estimates were weighted to represent a national sample. We analyzed PSI trends, Center for Medicaid and Medicare Services payment policy changes, and Inpatient Prospective Payment System regulations and notices between 2000 and 2013.

Results: Of the 13 PSIs studied, 10 had an overall decrease in rates and 3 had an increase. Joinpoint analysis showed that 12 of 13 PSIs had decreasing or stable trends in the last 5 years of the study. Central-line blood stream infections had the greatest annual decrease (-31.1 annual percent change between 2006 and 2013), whereas postoperative respiratory failure had the smallest decrease (-3.5 annual percent change between 2005 and 2013). With the exception of postoperative hip fracture, significant decreases in trends preceded federal payment reform initiatives.

Conclusions: National in-hospital patient safety has significantly improved between 2000 and 2015, as measured by PSIs. In this study, improvements in PSI trends often proceeded policies targeting patient safety events, suggesting that intense public discourses targeting patient safety may drive national policy reforms and that these improved trends may be sustained by the Center for Medicare and Medicaid Services policies that followed.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1097/PTS.0000000000000615DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8194008PMC
June 2021

Clinical Documentation to Predict Factors Associated with Urinary Incontinence Following Prostatectomy for Prostate Cancer.

Res Rep Urol 2020 23;12:7-14. Epub 2020 Jan 23.

Department of Medicine (Biomedical Informatics), Biomedical Data Sciences, and Surgery, Stanford University School of Medicine, Stanford, CA, USA.

Background: Advances in data collection provide opportunities to use population samples in identifying risk factors for urinary incontinence (UI), which occurs in up to 71% of men with prostate cancer following prostatectomy. Most studies on patient-centered outcomes use surveys or manual chart abstraction for data collection, which can be costly and difficult to scale. We sought to evaluate rates of and risk factors for UI following prostatectomy using natural language processing on electronic health record (EHR) data.

Methods: We conducted a retrospective analysis of patients undergoing prostatectomy for prostate cancer between January 2008 and August 2018 using EHR data from an academic medical center. UI incidence for each patient in the cohort was assessed using natural language processing from clinical notes generated pre- and postoperatively. Multivariable logistic regression was used to evaluate potential risk factors for postoperative UI at various time points within 2 years following surgery.

Results: We identified 3792 patients who underwent prostatectomy for prostate cancer. We found a significant association between preoperative UI and UI in the first (odds ratio [OR], 2.30; 95% confidence interval [CI], 1.24-4.28) and second (OR 2.24, 95% CI 1.04-4.83) years following surgery. Preoperative body mass index was also associated with UI in the second postoperative year (OR 1.11, 95% CI 1.02-1.21).

Conclusion: We show that a natural language processing approach using clinical narratives can be used to assess risk for UI in prostate cancer patients. Unstructured clinical narrative text can help advance future population-level research in patient-centered outcomes and quality of care.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.2147/RRU.S234178DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6986242PMC
January 2020

Leveraging Digital Data to Inform and Improve Quality Cancer Care.

Cancer Epidemiol Biomarkers Prev 2020 04 17;29(4):816-822. Epub 2020 Feb 17.

Stanford Cancer Institute, Stanford University School of Medicine, Stanford, California.

Background: Efficient capture of routine clinical care and patient outcomes is needed at a population-level, as is evidence on important treatment-related side effects and their effect on well-being and clinical outcomes. The increasing availability of electronic health records (EHR) offers new opportunities to generate population-level patient-centered evidence on oncologic care that can better guide treatment decisions and patient-valued care.

Methods: This study includes patients seeking care at an academic medical center, 2008 to 2018. Digital data sources are combined to address missingness, inaccuracy, and noise common to EHR data. Clinical concepts were identified and extracted from EHR unstructured data using natural language processing (NLP) and machine/deep learning techniques. All models are trained, tested, and validated on independent data samples using standard metrics.

Results: We provide use cases for using EHR data to assess guideline adherence and quality measurements among patients with cancer. Pretreatment assessment was evaluated by guideline adherence and quality metrics for cancer staging metrics. Our studies in perioperative quality focused on medications administered and guideline adherence. Patient outcomes included treatment-related side effects and patient-reported outcomes.

Conclusions: Advanced technologies applied to EHRs present opportunities to advance population-level quality assessment, to learn from routinely collected clinical data for personalized treatment guidelines, and to augment epidemiologic and population health studies. The effective use of digital data can inform patient-valued care, quality initiatives, and policy guidelines.

Impact: A comprehensive set of health data analyzed with advanced technologies results in a unique resource that facilitates wide-ranging, innovative, and impactful research on prostate cancer. This work demonstrates new ways to use the EHRs and technology to advance epidemiologic studies and benefit oncologic care.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1158/1055-9965.EPI-19-0873DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7195903PMC
April 2020

The rise of non-traumatic extremity compartment syndrome in light of the opioid epidemic.

Am J Emerg Med 2021 01 10;39:75-79. Epub 2020 Jan 10.

Division of Plastic & Reconstructive Surgery, Department of Surgery, Stanford University, United States of America; Department of Surgery, Palo Alto Veterans Hospital, United States of America. Electronic address:

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.ajem.2020.01.020DOI Listing
January 2021

The Impact of Hospital Quality on Thyroid Cancer Survival.

Otolaryngol Head Neck Surg 2020 03 21;162(3):269-276. Epub 2020 Jan 21.

Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California, USA.

Objective: To develop a composite measure of thyroid cancer-specific hospital quality and to evaluate the association between hospital quality and survival in patients with well-differentiated thyroid cancer.

Study Design: Retrospective cohort study.

Setting: Population-based cancer database.

Subjects And Methods: Data were extracted from the California Cancer Registry data set linked with discharge records and hospital characteristics from the California Office of Statewide Health Planning and Development. The study cohort comprised adult patients with well-differentiated thyroid cancer diagnosed between January 1, 2004, and December 31, 2015. Principal component analysis, incorporating hospital volume, adherence to national guidelines, and accreditation/certification status, was used to generate a composite thyroid cancer-specific hospital quality score.

Results: Treatment in hospitals ranked in the highest quartile of quality was associated with improved overall survival (OS) (hazard ratio [HR], 0.81; 95% confidence interval [CI], 0.67-0.98) and disease-specific survival (DSS) (HR, 0.72; 95% CI, 0.54-0.98). Treatment in hospitals meeting the combined metric of 10 or more thyroid cancer cases/year and 80% of patients with high-risk tumors treated with total/near-total thyroidectomy was associated with improved OS (HR, 0.80; 95% CI, 0.70-0.90) and DSS (HR, 0.77; 95% CI, 0.64-0.94).

Conclusion: Treatment in high-quality hospitals is associated with improved survival outcomes in patients with thyroid cancer. These findings are important because they help identify hospitals that are better suited to treat patients with thyroid cancer and provide actionable targets for quality improvement.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/0194599819900760DOI Listing
March 2020

Automated extraction of ophthalmic surgery outcomes from the electronic health record.

Int J Med Inform 2020 01 17;133:104007. Epub 2019 Oct 17.

Center for Biomedical Informatics Research, Stanford University, 1265 Welch Rd, Stanford, USA.

Objective: Comprehensive analysis of ophthalmic surgical outcomes is often restricted by limited methodologies for efficiently and accurately extracting clinical information from electronic health record (EHR) systems because much is in free-text form. This study aims to utilize advanced methods to automate extraction of clinical concepts from the EHR free text to study visual acuity (VA), intraocular pressure (IOP), and medication outcomes of cataract and glaucoma surgeries.

Methods: Patients who underwent cataract or glaucoma surgery at an academic medical center between 2009 and 2018 were identified by Current Procedural Terminology codes. Rule-based algorithms were developed and used on EHR clinical narrative text to extract intraocular lens (IOL) power and implant type, as well as to create a surgery laterality classifier. MedEx (version 1.3.7) was used on free-text clinical notes to extract information on eye medications and compared to information from medication orders. Random samples of free-text notes were reviewed by two independent masked annotators to assess inter-annotator agreement on outcome variable classification and accuracy of classifiers. VA and IOP were available from semi-structured fields.

Results: This study cohort included 6347 unique patients, with 8550 stand-alone cataract surgeries, 451 combined cataract/glaucoma surgeries, and 961 glaucoma surgeries without concurrent cataract surgery. The rule-based laterality classifier achieved 100% accuracy compared to manual review of a sample of operative notes by independent masked annotators. For cataract surgery alone, glaucoma surgery alone, or combined cataract/glaucoma surgeries, our automated extraction algorithm achieved 99-100% accuracy compared to manual annotation of samples of notes from each group, including IOL model and IOL power for cataract surgeries, and glaucoma implant for glaucoma surgeries. For glaucoma medications, there was 90.7% inter-annotator agreement. After adjudication, 85.0% of medications identified by MedEx determined to be correct. Determination of surgical laterality enabled evaluation of pre- and postoperative VA and IOP for operative eyes.

Conclusion: This text-processing pipeline can accurately capture surgical laterality and implant model usage from free-text operative notes of cataract and glaucoma surgeries, enabling extraction of clinical outcomes including visual acuities, intraocular pressure, and medications from the EHR system. Use of this approach with EHRs to assess ophthalmic surgical outcomes can benefit research groups interested in studying the safety and clinical efficacies of different surgical approaches.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.ijmedinf.2019.104007DOI Listing
January 2020

Trajectory analysis for postoperative pain using electronic health records: A nonparametric method with robust linear regression and K-medians cluster analysis.

Health Informatics J 2020 06 17;26(2):1404-1418. Epub 2019 Oct 17.

Stanford University, USA.

Postoperative pain scores are widely monitored and collected in the electronic health record, yet current methods fail to fully leverage the data with fast implementation. A robust linear regression was fitted to describe the association between the log-scaled pain score and time from discharge after total knee replacement. The estimated trajectories were used for a subsequent K-medians cluster analysis to categorize the longitudinal pain score patterns into distinct clusters. For each cluster, a mixture regression model estimated the association between pain score and time to discharge adjusting for confounding. The fitted regression model generated the pain trajectory pattern for given cluster. Finally, regression analyses examined the association between pain trajectories and patient outcomes. A total of 3442 surgeries were identified with a median of 22 pain scores at an academic hospital during 2009-2016. Four pain trajectory patterns were identified and one was associated with higher rates of outcomes. In conclusion, we described a novel approach with fast implementation to model patients' pain experience using electronic health records. In the era of big data science, clinical research should be learning from all available data regarding a patient's episode of care instead of focusing on the "average" patient outcomes.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1177/1460458219881339DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8012003PMC
June 2020
-->