SurvdigitizeR: an algorithm for automated survival curve digitization.

BACKGROUND: Decision analytic models and meta-analyses often rely on survival probabilities that are digitized from published Kaplan-Meier (KM) curves. However, manually extracting these probabilities from KM curves is time-consuming, expensive, and error-prone. We developed an efficient and accurate algorithm that automates extraction of survival probabilities from KM curves. METHODS: The automated digitization algorithm processes images from a JPG or PNG format, converts them in their hue, saturation, and lightness scale and uses optical character recognition to detect axis location and labels. It also uses a k-medoids clustering algorithm to separate multiple overlapping curves on the same figure. To validate performance, we generated survival plots form random time-to-event data from a sample size of 25, 50, 150, and 250, 1000 individuals split into 1,2, or 3 treatment arms. We assumed an exponential distribution and applied random censoring. We compared automated digitization and manual digitization performed by well-trained researchers. We calculated the root mean squared error (RMSE) at 100-time points for both methods. The algorithm's performance was also evaluated by Bland-Altman analysis for the agreement between automated and manual digitization on a real-world set of published KM curves. RESULTS: The automated digitizer accurately identified survival probabilities over time in the simulated KM curves. The average RMSE for automated digitization was 0.012, while manual digitization had an average RMSE of 0.014. Its performance was negatively correlated with the number of curves in a figure and the presence of censoring markers. In real-world scenarios, automated digitization and manual digitization showed very close agreement. CONCLUSIONS: The algorithm streamlines the digitization process and requires minimal user input. It effectively digitized KM curves in simulated and real-world scenarios, demonstrating accuracy comparable to conventional manual digitization. The algorithm has been developed as an open-source R package and as a Shiny application and is available on GitHub: https://github.com/Pechli-Lab/SurvdigitizeR and https://pechlilab.shinyapps.io/SurvdigitizeR/ .

Local practice variations and payer differences underlie state-wide disparities in oncoplastic breast surgery.

BACKGROUND: Little is known about disparities in oncoplastic breast surgery delivery. METHODS: The Massachusetts All-Payer Claims Database was queried for patients who received lumpectomy for a diagnosis of breast cancer. Oncoplastic surgery was defined as adjacent tissue transfer, complex trunk repair, reduction mammoplasty, mastopexy, flap-based reconstruction, prosthesis insertion, or unspecified breast reconstruction after lumpectomy. RESULTS: We identified 18 748 patients who underwent lumpectomy between 2016 and 2020. Among those, 3140 patients underwent immediate oncoplastic surgery and 436 patients underwent delayed oncoplastic surgery. Eighty-one percent of patients who underwent oncoplastic surgery did so in the same county as they underwent a lumpectomy. However, the relative frequency of oncoplastic surgery varied significantly among counties. In multivariable regression, public insurance status (odds ratio: 0.87, 95% confidence interval: 0.80-0.95, p = 0.002) was associated with lower odds of undergoing oncoplastic surgery, even after adjusting for macromastia, other comorbidities, and county of lumpectomy. Average payments for lumpectomy with oncoplastic surgery were more than twice as high from private insurers ($840 vs. $1942, p < 0.001). CONCLUSION: Disparities in the receipt of oncoplastic surgery were related to differences in local practice patterns and the type of insurance patients held. Expanding services across counties and considering billing reform may help reduce these disparities.

Geographic Disparities and Payment Variation for Immediate Lymphatic Reconstruction in Massachusetts.

BACKGROUND: Little is known about practice patterns and payments for immediate lymphatic reconstruction (ILR). This study aims to evaluate trends in ILR delivery and billing practices. METHODS: We queried the Massachusetts All-Payer Claims Database between 2016 and 2020 for patients who underwent lumpectomy or mastectomy with axillary lymph node dissection for oncologic indications. We further identified patients who underwent lymphovenous bypass on the same date as tumor resection. We used ZIP code data to analyze the geographic distribution of ILR procedures and calculated physician payments for these procedures, adjusting for inflation. We used multivariable logistic regression to identify variables, which predicted receipt of ILR. RESULTS: In total, 2862 patients underwent axillary lymph node dissection over the study period. Of these, 53 patients underwent ILR. Patients who underwent ILR were younger (55.1 vs 59.3 years, P = 0.023). There were no significant differences in obesity, diabetes, or smoking history between the two groups. A greater percentage of patients who underwent ILR had radiation (83% vs 67%, P = 0.027). In multivariable regression, patients residing in a county neighboring Boston had 3.32-fold higher odds of undergoing ILR (95% confidence interval: 1.76-6.25; P < 0.001), while obesity, radiation therapy, and taxane-based chemotherapy were not significant predictors. Payments for ILR varied widely. CONCLUSIONS: In Massachusetts, patients were more likely to undergo ILR if they resided near Boston. Thus, many patients with the highest known risk for breast cancer-related lymphedema may face barriers accessing ILR. Greater awareness about referring high-risk patients to plastic surgeons is needed.

Outcomes of Dermal Regeneration Templates in Irradiated and Nonirradiated Scalp Defects.

OBJECTIVES: Dermal regeneration templates (DRTs) are frequently used to treat scalp defects. The aim was to compare the time course of healing for DRTs in scalp defects with and without preoperative radiation. METHODS: The authors conducted a retrospective cohort study of DRT-based scalp reconstruction at 2 academic medical centers between 2013 and 2022. Information was collected on demographic variables, comorbidities, medication use, history of radiation, and DRT outcomes. The primary outcome was DRT loss, defined as exposed calvarium or DRT detachment based on postoperative follow-up documentation. Kaplan-Meier survival analysis and multivariable Cox proportional-hazard regressions were used to compare DRT loss in irradiated and nonirradiated defects. Multivariable logistic regressions were used to compare 30-day postoperative complications (infection, hematoma, or seroma) in irradiated and nonirradiated defects. RESULTS: In total, 158 cases were included. Twenty-eight (18%) patients had a preoperative history of radiation to the scalp. The mean follow-up time after DRT placement was 2.6 months (SD: 4.5 mo). The estimated probability of DRT survival at 2 months was 91% (95% CI: 83%-100%) in nonirradiated patients and 65% (95% CI: 48%-88%) in irradiated patients. In the 55 patients with a bony wound base, preoperative head radiation was associated with a higher likelihood of DRT loss (hazard ratio: 11). Half the irradiated defects experienced uncomplicated total wound closure using Integra Wound Matrix Dressing with or without second-stage reconstruction. CONCLUSIONS: Dermal regeneration template can offer durable coverage in nonirradiated scalp defects. Although DRT loss is more likely in irradiated scalp defects, successful DRT-based reconstruction is possible in select cases.

The Use of Unlisted Billing Codes for Microsurgical Breast Reconstruction and Implications for Code Consolidation.

BACKGROUND: Private insurers have considered consolidating the billing codes presently available for microvascular breast reconstruction. There is a need to understand how these different codes are currently distributed and used to help inform how coding consolidation may impact patients and providers. METHODS: Using the Massachusetts All-Payer Claims Database between 2016 and 2020, patients who underwent microsurgical breast reconstruction following mastectomy for cancer-related indications were identified. Multivariable logistic regression was used to test whether an S2068 claim was associated with insurance type and median household income by patient ZIP code. The ratio of S2068 to CPT19364 claims for privately insured patients was calculated for providers practicing in each county. Total payments for professional fees were compared between billing codes. RESULTS: There were 272 claims for S2068 and 209 claims for CPT19364. An S2068 claim was associated with age < 45 years (OR: 1.89, 95% CI: 1.11-3.20, p = 0.019), more affluent ZIP codes (OR: 1.11, 95% CI: 1.03-1.19, p = 0.004), and private insurance (OR: 16.13, 95% CI: 7.81-33.33, p < 0.001). Median total payments from private insurers were 101% higher for S2068 than for CPT19364. In all but two counties (Worcester and Hampshire), the S-code was used more frequently than CPT19364 for their privately insured patients. CONCLUSION: Coding practices for microsurgical breast reconstruction lacked uniformity in Massachusetts, and payments differed greatly between S2068 and CPT19364. Patients from more affluent towns were more likely to have S-code claims. Coding consolidation could impact access, as the majority of providers in Massachusetts might need to adapt their practices if the S-code were discontinued.

Rates of price disclosure associated with the surgical treatment of early-stage breast cancer one year after implementation of federal regulations.

PURPOSE: Mastectomy, breast reconstruction (BR) and breast conserving therapy (BCT) are core components of the treatment paradigm for early-stage disease but are differentially associated with significant financial burdens. Given recent price transparency regulations, we sought to characterize rates of disclosure for breast cancer-related surgery, including mastectomy, BCT, and BR (oncoplastic reconstruction, implant, pedicled flap and free flap) and identify associated factors. METHODS: For this cross-sectional analysis, cost reports were obtained from the Turquoise Health price transparency platform for all U.S. hospitals meeting national accreditation standards for breast cancer care. The Healthcare Cost Report Information System was used to collect facility-specific data. Addresses were geocoded to identify hospital referral and census regions while data from CMS was also used to identify the geographic practice cost index. We leveraged a Poisson regression model and relevant Medicare billing codes to analyze factors associated with price disclosure and the availability of an OOP price estimator. RESULTS: Of 447 identified hospitals, 221 (49.4%) disclosed prices for mastectomy and 188 42.1%) disclosed prices for both mastectomy and some form of reconstruction including oncoplastic reduction (n = 184, 97.9%), implants (n = 187, 99.5%), pedicled flaps (n = 89, 47.3%), and free flaps (n = 81, 43.1%). Non-profit status and increased market competition were associated with price nondisclosure. 121 hospitals (27.1%) had an out-of-pocket price estimator that included at least one breast surgery. CONCLUSIONS: Most eligible hospitals did not disclose prices for breast cancer surgery. Distinct hospital characteristics were associated with price disclosure. Breast cancer patients face persistent difficulty in accessing costs.

Having Social Media among Integrated Plastic Surgery Applicants: Is It Needed to Match?

Background: Little information exists on the perceptions of integrated plastic and reconstructive surgery (PRS) residency applicants on the need for having social media (SoMe) during the application process. Methods: A cross-sectional survey study was conducted during the 2022 match cycle to assess integrated PRS residency applicants' perceptions on the role of SoMe during the match. Univariate and multivariate analyses were performed on variables of interest. Qualitative analysis was conducted on free-form responses. Results: Seventy-nine surveys were completed (response rate: 24%). The majority of respondents were educated in the United States (92%). Instagram was the most commonly used SoMe platform (92%). Of those surveyed, 18% thought that SoMe was beneficial to the application process. Twenty-nine percent of respondents agreed that a SoMe presence increases one's chances of matching into PRS residency (41% disagreed and 30% responded neutrally). Forty-four percent endorsed stress about maintaining a SoMe presence in PRS. Having mentors who recommended maintaining a SoMe presence was associated with the belief that SoMe increases one's chances of matching [odds ratio (OR) 8.1, 95% confidence interval (CI) 1.1-40.4, P = 0.011] and stress about maintaining a SoMe presence (OR 6.3, 95% CI 1.2-33.3, P = 0.030). Applicants who did research years had lower odds of experiencing stress (OR 0.16, 95% CI 0.04-0.70, P = 0.015). Conclusions: The growing role of SoMe in the residency selection process may be exacerbating applicants' stress and anxiety. PRS programs may consider establishing clear policies for how SoMe will be used in evaluating candidates.

In-person, Hybrid, and Virtual Grand Rounds in Plastic Surgery: Impact and Participant Perceptions.

Many plastic surgery residency programs adapted to the COVID-19 pandemic by implementing virtual grand rounds. This study aimed to assess the impact of virtual grand rounds and how attendees perceived virtual grand rounds to inform future programmatic planning. Methods: This was a quality improvement initiative involving a cross-sectional survey and retrospective review of administrative records for the 2017-2018 (in-person) and 2021-2022 (virtual) academic years for two academic plastic surgery training programs in Boston, MA. Respondents were residents, fellows, and faculty within the two multisite plastic surgery residency training programs. Results: There were 39 respondents (51% faculty, 41% residents, and 8% fellows). There was no evidence of different preferences for the format of future grand rounds (P = 0.08), with most preferring hybrid, defined as in person for speakers and others who could attend. Most respondents indicated a more accessible learning environment (86.8%) and lack of in-person interaction (82.1%) as reasons for liking and not liking virtual grand rounds, respectively. Excluding outliers, attendance in 2021-2022 was on average 7.4% points greater than that in 2017-2018 (P < 0.001), or six to seven more individuals at each session. There were significantly more out-of-state speakers in 2021-2022 (84%) as compared to 2017-2018 (28%) (P = 0.0008). Conclusions: Virtual grand rounds improved attendance and the geographic diversity of speakers. Attendees preferred a hybrid format for future grand rounds, citing advantages and disadvantages to both in-person and virtual formats.

Healthcare Value in Plastic Surgery Practice.

BRIEF SUMMARY: As value-based care gains traction in response to towering healthcare expenditures and issues of healthcare inequity, hospital capacity, and labor shortages, it is important to consider how a value-based approach can be achieved in plastic surgery. Value is defined as outcomes divided by costs across entire cycles of care. Drawing on previous studies and policies, this paper identifies key opportunities in plastic surgery to move the levers of costs and outcomes to deliver higher-value care. Specifically, outcomes in plastic surgery should include conventional measures of complication rates as well as patient-reported outcome measures in order to drive quality improvement and benchmark payments. Meanwhile, cost reduction in plastic surgery can be achieved through value-based payment reform, efficient workflows, evidence-based and cost-conscious selection of medical devices, and greater use of out-patient surgical facilities. Lastly, we discuss how the diminished presence of third-party payers in aesthetic surgery exemplifies the cost-conscious and patient-centered nature of value-based plastic surgery. To lead in future health policy and care delivery reform, plastic surgeons should strive for high-value care, remain open to new ways of care delivery, and understand how plastic surgery fits into overall health care delivery.

Uncinate Duct Dilatation Predicts Additional Risk for High-Grade Dysplasia or Invasive Carcinoma Among Fukuoka-Positive Intraductal Papillary Mucinous Neoplasms.

OBJECTIVE: To determine whether uncinate duct dilatation (UDD) increases the risk of high-grade dysplasia or invasive carcinoma (HGD/IC) in Fukuoka-positive intraductal papillary mucinous neoplasms (IPMNs). BACKGROUND: Though classified as a branch duct, the uncinate duct is the primary duct of the pancreatic ventral anlage. We hypothesized that UDD, like main duct dilatation, confers additional risk for HGD/IC. METHODS: A total of 467 patients met inclusion criteria in a retrospective cohort study of surgically resected IPMNs at the Massachusetts General Hospital. We used multivariable logistic regression to analyze the association between UDD (defined as ≥4 mm) and HGD/IC, controlling for Fukuoka risk criteria. In a secondary analysis, the modeling was repeated in the 194 patients with dorsal branch duct IPMNs (BD-IPMNs) in the pancreatic neck, body, or tail. RESULTS: Mean age at surgery was 70, and 229 (49%) patients were female. In total, 267 (57%) patients had only worrisome features and 200 (43%) had at least 1 high-risk feature. UDD was present in 164 (35%) patients, of whom 118 (73%) had HGD/IC. On multivariable analysis, UDD increased the odds of HGD/IC by 2.8-fold, even while controlling for Fukuoka risk factors (95% CI: 1.8-4.4, P <0.001). Prevalence of HGD/IC in all patients with UDD was 73%, compared with 74% in patients with high-risk stigmata and 73% in patients with main duct IPMNs. In the secondary analysis, UDD increased the odds of HGD/IC by 3.2-fold in patients with dorsal BD-IPMNs (95% CI: 1.3-7.7, P =0.010). CONCLUSIONS: UDD confers additional risk for HGD/IC unaccounted for by current Fukuoka criteria. Further research can extend this study to Fukuoka-negative patients, including unresected patients.

Black-versus-White racial disparities in 30-day outcomes at Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program-accredited centers: a needed quality indicator.

BACKGROUND: Creating a metric in the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) to assess Black-versus-White disparities is critical if we are to ensure equitable care for all. OBJECTIVE: To investigate Black-versus-White disparities while replicating MBSAQIP methodology with regard to covariates and modeling so that the results can serve as the foundation to create a benchmarked site-level Disparities Metric for MBSAQIP. SETTING: United States and Canada. METHODS: Across the 2015-2019 MBSAQIP cohorts, 543,976 adults underwent primary or revision sleeve gastrectomy or Roux-en-Y gastric bypass and were reported as either White or Black. Using a set of covariates derived from published MBSAQIP performance models, we performed multivariable logistic modeling with 10-fold cross-validation for the 11 outcomes evaluated in MBSAQIP Semiannual Reports, plus venous thromboembolism (VTE) and death. We analyzed primary and revision cases separately. RESULTS: After risk adjustment, Black patients experienced higher odds of all-occurrence morbidity (odds ratio [OR], 1.22; 95% confidence interval [CI], 1.19-1.25; P < .001), serious events (OR, 1.08; 95% CI, 1.04-1.13; P < .001), all-cause intervention (OR, 1.31; 95% CI, 1.24-1.37; P < .001), related intervention (OR, 1.29; 95% CI, 1.22-1.37; P < .001), all-cause readmission (OR, 1.37; 95% CI, 1.33-1.41; P < .001), related readmission (OR, 1.41; 95% CI, 1.36-1.46; P < .001), venous thromboembolism (OR, 1.49; 95% CI, 1.34-1.65; P < .001), and death (OR, 1.59; 95% CI, 1.34-1.89; P < .001) after primary procedures. Black patients experienced lower odds of morbidity (OR, .94; 95% CI, .91-.98; P = .004) and surgical-site infection (OR, .72; 95% CI, .66-.78; P < .001). CONCLUSIONS: Black patients experienced a higher risk for serious complications and required more readmissions, reoperations, and postoperative interventions. This study supports the creation of a site-level Disparities Metric for the MBSAQIP and provides the framework to do so.

An Introductory Tutorial on Cohort State-Transition Models in R Using a Cost-Effectiveness Analysis Example.

Decision models can combine information from different sources to simulate the long-term consequences of alternative strategies in the presence of uncertainty. A cohort state-transition model (cSTM) is a decision model commonly used in medical decision making to simulate the transitions of a hypothetical cohort among various health states over time. This tutorial focuses on time-independent cSTM, in which transition probabilities among health states remain constant over time. We implement time-independent cSTM in R, an open-source mathematical and statistical programming language. We illustrate time-independent cSTMs using a previously published decision model, calculate costs and effectiveness outcomes, and conduct a cost-effectiveness analysis of multiple strategies, including a probabilistic sensitivity analysis. We provide open-source code in R to facilitate wider adoption. In a second, more advanced tutorial, we illustrate time-dependent cSTMs.

A Tutorial on Time-Dependent Cohort State-Transition Models in R Using a Cost-Effectiveness Analysis Example.

In an introductory tutorial, we illustrated building cohort state-transition models (cSTMs) in R, where the state transition probabilities were constant over time. However, in practice, many cSTMs require transitions, rewards, or both to vary over time (time dependent). This tutorial illustrates adding 2 types of time dependence using a previously published cost-effectiveness analysis of multiple strategies as an example. The first is simulation-time dependence, which allows for the transition probabilities to vary as a function of time as measured since the start of the simulation (e.g., varying probability of death as the cohort ages). The second is state-residence time dependence, allowing for history by tracking the time spent in any particular health state using tunnel states. We use these time-dependent cSTMs to conduct cost-effectiveness and probabilistic sensitivity analyses. We also obtain various epidemiological outcomes of interest from the outputs generated from the cSTM, such as survival probability and disease prevalence, often used for model calibration and validation. We present the mathematical notation first, followed by the R code to execute the calculations. The full R code is provided in a public code repository for broader implementation.

Mobile Health Interventions and RCTs: Structured Taxonomy and Research Framework.

Mobile Health Interventions (MHIs) have addressed a range of healthcare challenges and have been evaluated using Randomized Controlled Trials (RCTs) to establish clinical effectiveness. Using PRISMA we conducted a systematic literature review of RCTs for MHIs and identified 70 studies which were analyzed and classified using Nickerson-Varshney-Muntermann (NVM) taxonomy. From the resultant iterations of the taxonomy, we extracted insights from the categorized studies. RCTs cover a wide range of health conditions including chronic diseases, general wellness, unhealthy practices, family planning, end-of-life, and post-transplant care. The MHIs that were utilized by the RCTs were varied as well, although most studies did not find significant differences between MHIs and usual care. The challenges for MHI-based RCTs include the use of technologies, delayed outcomes, patient recruitment, patient retention, and complex regulatory requirements. These variances can lead to a higher rate of Type I/Type II errors. Further considerations are the impact of infrastructure, contextual and cultural factors, and reductions in the technological relevancy of the intervention itself. Finally, due to the delayed effect of most outcomes, RCTs of insufficient duration are unable to measure significant, lasting improvements. Using the insights from seventy identified studies, we developed a classification of existing RCTs along with guidelines for MHI-based RCTs and a research framework for future RCTs. The framework offers opportunities for (a) personalization of MHIs, (b) use of richer technologies, and (c) emerging areas for RCTs.

Environmental variables and genome-environment interactions predicting IBD diagnosis in large UK cohort.

A combination of genetic susceptibility and environmental exposure is thought to cause inflammatory bowel disease (IBD), but the non-genetic component remains poorly characterized. We therefore undertook a search for environmental variables and gene-environment interactions associated with future IBD diagnosis in a large UK cohort. Using self-report and electronic health records, we identified 1946 Crohn's disease (CD) and 3715 ulcerative colitis (UC) patients after quality control in the UK Biobank. Based on prior literature and biological plausibility , we tested 38 candidate environmental variables for association with CD, UC, and overall IBD using Cox proportional hazard regressions. We also tested whether these variables interacted with polygenic risk in predicting disease, following up significant (FDR < 0.05) results with tests for SNP-environment associations. We performed robustness analyses on all significant results. As in previous reports, appendectomy protected against UC, smoking (both current and previous) elevated risk for CD, current smoking protected against UC, and previous smoking imparted a risk for UC. Childhood antibiotic use associated with IBD, as did sun exposure during the winter. Socioeconomic deprivation was conferred a risk for IBD, CD, and UC. We uncovered negative interactions between polygenic risk and previous oral contraceptive use for IBD and UC. Polygenic risk also interacted negatively with previous smoking in predicting UC. There were no individually significant SNP-environment interactions. Thus, for a limited set of environmental variables, there was strong evidence of association with IBD diagnosis in the UK Biobank, and interaction with polygenic risk was minimal.

Interpreting and assessing confidence in network meta-analysis results: an introduction for clinicians.

PURPOSE: We aimed to provide clinicians with introductory guidance for interpreting and assessing confidence in on Network meta-analysis (NMA) results. METHODS: We reviewed current literature on NMA and summarized key points. RESULTS: Network meta-analysis (NMA) is a statistical method for comparing the efficacy of three or more interventions simultaneously in a single analysis by synthesizing both direct and indirect evidence across a network of randomized clinical trials. It has become increasingly popular in healthcare, since direct evidence (head-to-head randomized clinical trials) are not always available. NMA methods are categorized as either Bayesian or frequentist, and while the two mostly provide similar results, the two approaches are theoretically different and require different interpretations of the results. CONCLUSIONS: We recommend a careful approach to interpreting NMA results and the validity of an NMA depends on its underlying statistical assumptions and the quality of the evidence used in the NMA.