An Electronic Health Record Text Mining Tool to Collect Real-World Drug Treatment Outcomes: A Validation Study in Patients With Metastatic Renal Cell Carcinoma

Automated derivation of diagnostic criteria for lung cancer using natural language processing on electronic health records: a pilot study

BACKGROUND

The digitisation of healthcare records has generated vast amounts of unstructured data, presenting opportunities for improvements in disease diagnosis when clinical coding falls short, such as in the recording of patient symptoms. This study presents an approach using natural language processing to extract clinical concepts from free-text which are used to automatically form diagnostic criteria for lung cancer from unstructured secondary-care data.

METHODS

Patients aged 40 and above who underwent a chest x-ray (CXR) between 2016 and 2022 were included. ICD-10 and unstructured data were pulled from their electronic health records (EHRs) over the preceding 12 months to the CXR. The unstructured data were processed using named entity recognition to extract symptoms, which were mapped to SNOMED-CT codes. Subsumption of features up the SNOMED-CT hierarchy was used to mitigate against sparse features and a frequency-based criteria, combined with univariate logarithmic probabilities, was applied to select candidate features to take forward to the model development phase. A genetic algorithm was employed to identify the most discriminating features to form the diagnostic criteria.

RESULTS

75002 patients were included, with 1012 lung cancer diagnoses made within 12 months of the CXR. The best-performing model achieved an AUROC of 0.72. Results showed that an existing 'disorder of the lung', such as pneumonia, and a 'cough' increased the probability of a lung cancer diagnosis. 'Anomalies of great vessel', 'disorder of the retroperitoneal compartment' and 'context-dependent findings', such as pain, statistically reduced the risk of lung cancer, making other diagnoses more likely. The performance of the developed model was compared to the existing cancer risk scores, demonstrating superior performance.

CONCLUSIONS

The proposed methods demonstrated success in leveraging unstructured secondary-care data to derive diagnostic criteria for lung cancer, outperforming existing risk tools. These advancements show potential for enhancing patient care and results. However, it is essential to tackle specific limitations by integrating primary care data to ensure a more thorough and unbiased development of diagnostic criteria. Moreover, the study highlights the importance of contextualising SNOMED-CT concepts into meaningful terminology that resonates with clinicians, facilitating a clearer and more tangible understanding of the criteria applied.

Riboflavin transporter deficiency, the search for the undiagnosed: a retrospective data mining study

BACKGROUND

Riboflavin transporter deficiency (RTD) is an inborn error of riboflavin transport causing progressive neurological symptoms if left untreated. While infants with symptomatic RTD rapidly deteriorate, presentation later in childhood or in adulthood is more gradual. Symptoms overlap with more common diseases, carrying a risk of misdiagnosis, and given the relatively recent discovery of the genetic basis of RTD in 2010 it is likely that older patients have not been tested. Treatment with oral riboflavin (vitamin B2) halts disease progression and can be lifesaving. We hypothesized that patients may have been left unrecognized at the time of presentation and therefore we performed a datamining study to detect undiagnosed RTD patients in a tertiary referral hospital.

METHODS

A systematic search in Electronic Health Records (EHR) of all patients visiting the Amsterdam University Medical Centers between January 2004 and July 2021 was performed by a medical data text-mining tool. Pseudonymized patient records, matching pre-defined search terms (hearing loss or auditory neuropathy spectrum disorders combined with key clinical symptoms or riboflavin) were screened and included if no definitive alternative diagnosis for symptoms indicating possible RTD was found. Included patients were offered genetic testing. We documented total number of patients with possible RTD, number of patients that underwent genetic testing for RTD and results of genetic testing.

RESULTS

EHR of 2.288.901 patients were automatically screened. Thirteen patients with possible RTD were identified and offered genetic testing. Seven patients chose not to participate. Genetic testing was performed in 6 patients and was negative. The datamining did detect all previously known RTD patients in the hospital.

CONCLUSIONS

By screening a large cohort of patients of all ages in a tertiary referral hospital in a period spanning 17 years, no new RTD patients were found. Although not all suspected patients underwent genetic testing, our findings suggest that the prevalence of RTD is low and the chance of having missed this diagnosis in a tertiary referral hospital is limited.

Development and Portability of a Text Mining Algorithm for Capturing Disease Progression in Electronic Health Records of Patients With Stage IV Non-Small Cell Lung Cancer

PURPOSE

The objective was to develop and evaluate the portability of a text mining algorithm for prospectively capturing disease progression in electronic health record (EHR) data of patients with metastatic non-small cell lung cancer (mNSCLC) treated with immunochemotherapy.

METHODS

This study used EHR data from patients with mNSCLC receiving immunochemotherapy (between October 1, 2018, and December 31, 2022) in four Dutch hospitals. A text mining algorithm for capturing disease progression was developed in hospitals 1 and 2 and then transferred to hospitals 3 and 4 to evaluate portability. Performance metrics were calculated by comparing its outcomes with manual chart review. In addition, data were simulated to come available over time to assess performance in real-time applications. Median progression-free survival (PFS) was calculated using the Kaplan-Meier method to compare text mining with manual chart review.

RESULTS

During development and portability, the text mining algorithm performed well in capturing disease progression, with all performance scores >90%. When real-time performance was simulated, the performance scores in all four hospitals exceeded 90% from week 15 after the start of follow-up. Although the exact progression dates varied in 46 patients of 157 patients with progressive disease, the number of patients labeled with progression too early (n = 24) and too late (n = 22) was well balanced with discrepancies ranging from -116 to 384 days. Nevertheless, the PFS curves constructed with text mining and manual chart review were highly similar for each hospital.

CONCLUSION

In this study, an accurate text mining algorithm for capturing disease progression in the EHR data of patients with mNSCLC was developed. The algorithm was portable across different hospitals, and the performance over time was good, making this an interesting approach for prospective follow-up of multicenter cohorts.

Women have a higher risk of hospital admission associated with hyponatremia than men while using diuretics

Background

Hyponatremia is a common electrolyte disturbance and known adverse drug reaction of diuretics. Women tend to be more susceptible for diuretic associated hyponatremia. The aim of this study was to find more evidence whether women have a higher risk of diuretic associated hyponatremia than men measured at hospital admission for specific diuretic groups and whether there is a sex difference in risk of severity of hyponatremia.

Methods

All patients using a diuretic and admitted for any reason to Tergooi MC and Haga Teaching hospital in the Netherlands between the 1st of January 2017 and the 31st of December 2021, with recorded sodium levels at admission were included in this study. Cases were defined as patients with a sodium level <135 mmol/L, while control patients had a sodium level ≥135 mmol/L at admission. Logistic regression analysis was used to calculate odds ratios (OR) with 95% CIs for women versus men and adjusted for potential confounding covariables (age, body mass index, potassium serum level, systolic and diastolic blood pressure, estimated glomerular filtration rate, number of diuretics, comedications and comorbidities). Stratified analyses were conducted for specific diuretic groups (thiazides, loop diuretics and aldosterone antagonists), and adjusted for dose. Furthermore, stratified analyses were performed by severity of hyponatremia (severe: <125 mmol/L), mild: 125-134 mmol/L).

Results

A total of 2,506 patients (50.0% women) were included, of which 516 had hyponatremia at admission (20.6%, 56.2% women). Women had a statistically significantly higher risk for hyponatremia at admission than men (OR 1.37; 95% CI 1.12-1.66) and after adjustment for potential risk factors (ORadj 1.55; 95% CI 1.22-1.98). Stratified analyses showed increased odds ratios for thiazides (ORadj 1.35; 95% CI 1.00-1.83) and loop diuretics (ORadj 1.62; 95% CI 1.19-2.19) among women. Use of aldosterone antagonists was also increased but not statistically significant (ORadj 1.15; 95% CI 0.73-1.81). Women had a statistically higher risk to develop mild and severe hyponatremia than men (ORadj 1.36; 95% CI 1.10-1.68 and ORadj 1.96; 95%CI 1.04-3.68, respectively).

Conclusion

Women have a higher risk of a hospital admission associated with hyponatremia while using diuretics than men. Further research is necessary to provide sex-specific recommendations.

Use of proxy indicators for automated surveillance of severe acute respiratory infection, the Netherlands, 2017 to 2023: a proof-of-concept study

BackgroundEffective pandemic preparedness requires robust severe acute respiratory infection (SARI) surveillance. However, identifying SARI patients based on symptoms is time-consuming. Using the number of reverse transcription (RT)-PCR tests or contact and droplet precaution labels as a proxy for SARI could accurately reflect the epidemiology of patients presenting with SARI.AimWe aimed to compare the number of RT-PCR tests, contact and droplet precaution labels and SARI-related International Classification of Disease (ICD)-10 codes and evaluate their use as surveillance indicators.MethodsPatients from all age groups hospitalised at Leiden University Medical Center between 1 January 2017 up to and including 30 April 2023 were eligible for inclusion. We used a clinical data collection tool to extract data from electronic medical records. For each surveillance indicator, we plotted the absolute count for each week, the incidence proportion per week and the correlation between the three surveillance indicators.ResultsWe included 117,404 hospital admissions. The three surveillance indicators generally followed a similar pattern before and during the COVID-19 pandemic. The correlation was highest between contact and droplet precaution labels and ICD-10 diagnostic codes (Pearson correlation coefficient: 0.84). There was a strong increase in the number of RT-PCR tests after the start of the COVID-19 pandemic.DiscussionAll three surveillance indicators have advantages and disadvantages. ICD-10 diagnostic codes are suitable but are subject to reporting delays. Contact and droplet precaution labels are a feasible option for automated SARI surveillance, since these reflect trends in SARI incidence and may be available real-time.

An innovative method to strengthen evidence for potential drug safety signals using Electronic Health Records

Reports from spontaneous reporting systems (SRS) are hypothesis generating. Additional evidence such as more reports is required to determine whether the generated drug-event associations are in fact safety signals. However, underreporting of adverse drug reactions (ADRs) delays signal detection. Through the use of natural language processing, different sources of real-world data can be used to proactively collect additional evidence for potential safety signals. This study aims to explore the feasibility of using Electronic Health Records (EHRs) to identify additional cases based on initial indications from spontaneous ADR reports, with the goal of strengthening the evidence base for potential safety signals. For two confirmed and two potential signals generated by the SRS of the Netherlands Pharmacovigilance Centre Lareb, targeted searches in the EHR of the Leiden University Medical Centre were performed using a text-mining based tool, CTcue. The search for additional cases was done by constructing and running queries in the structured and free-text fields of the EHRs. We identified at least five additional cases for the confirmed signals and one additional case for each potential safety signal. The majority of the identified cases for the confirmed signals were documented in the EHRs before signal detection by the Dutch Medicines Evaluation Board. The identified cases for the potential signals were reported to Lareb as further evidence for signal detection. Our findings highlight the feasibility of performing targeted searches in the EHR based on an underlying hypothesis to provide further evidence for signal generation.

Modest performance of text mining to extract health outcomes may be almost sufficient for high-quality prognostic model development

BACKGROUND

Across medicine, prognostic models are used to estimate patient risk of certain future health outcomes (e.g., cardiovascular or mortality risk). To develop (or train) prognostic models, historic patient-level training data is needed containing both the predictive factors (i.e., features) and the relevant health outcomes (i.e., labels). Sometimes, when the health outcomes are not recorded in structured data, these are first extracted from textual notes using text mining techniques. Because there exist many studies utilizing text mining to obtain outcome data for prognostic model development, our aim is to study the impact of the text mining quality on downstream prognostic model performance.

METHODS

We conducted a simulation study charting the relationship between text mining quality and prognostic model performance using an illustrative case study about in-hospital mortality prediction in intensive care unit patients. We repeatedly developed and evaluated a prognostic model for in-hospital mortality, using outcome data extracted by multiple text mining models of varying quality.

RESULTS

Interestingly, we found in our case study that a relatively low-quality text mining model (F1 score ≈ 0.50) could already be used to train a prognostic model with quite good discrimination (area under the receiver operating characteristic curve of around 0.80). The calibration of the risks estimated by the prognostic model seemed unreliable across the majority of settings, even when text mining models were of relatively high quality (F1 ≈ 0.80).

DISCUSSION

Developing prognostic models on text-extracted outcomes using imperfect text mining models seems promising. However, it is likely that prognostic models developed using this approach may not produce well-calibrated risk estimates, and require recalibration in (possibly a smaller amount of) manually extracted outcome data.

Using an artificial intelligence tool incorporating natural language processing to identify patients with a diagnosis of ANCA-associated vasculitis in electronic health records

BACKGROUND

Because anti-neutrophil cytoplasmatic antibody (ANCA)-associated vasculitis (AAV) is a rare, life-threatening, auto-immune disease, conducting research is difficult but essential. A long-lasting challenge is to identify rare AAV patients within the electronic-health-record (EHR)-system to facilitate real-world research. Artificial intelligence (AI)-search tools using natural language processing (NLP) for text-mining are increasingly postulated as a solution.

METHODS

We employed an AI-tool that combined text-mining with NLP-based exclusion, to accurately identify rare AAV patients within large EHR-systems (>2.000.000 records). We developed an identification method in an academic center with an established AAV-training set (n = 203) and validated the method in a non-academic center with an AAV-validation set (n = 84). To assess accuracy anonymized patient records were manually reviewed.

RESULTS

Based on an iterative process, a text-mining search was developed on disease description, laboratory measurements, medication and specialisms. In the training center, 608 patients were identified with a sensitivity of 97.0 % (95%CI [93.7, 98.9]) and positive predictive value (PPV) of 56.9 % (95%CI [52.9, 60.1]). NLP-based exclusion resulted in 444 patients increasing PPV to 77.9 % (95%CI [73.7, 81.7]) while sensitivity remained 96.3 % (95%CI [93.8, 98.0]). In the validation center, text-mining identified 333 patients (sensitivity 97.6 % (95%CI [91.6, 99.7]), PPV 58.2 % (95%CI [52.8, 63.6])) and NLP-based exclusion resulted in 223 patients, increasing PPV to 86.1 % (95%CI [80.9, 90.4]) with 98.0 % (95%CI [94.9, 99.4]) sensitivity. Our identification method outperformed ICD-10-coding predominantly in identifying MPO+ and organ-limited AAV patients.

CONCLUSIONS

Our study highlights the advantages of implementing AI, notably NLP, to accurately identify rare AAV patients within large EHR-systems and demonstrates the applicability and transportability. Therefore, this method can reduce efforts to identify AAV patients and accelerate real-world research, while avoiding bias by ICD-10-coding.

Reliability and Efficiency of the CAPRI-3 Metastatic Prostate Cancer Registry Driven by Artificial Intelligence

BACKGROUND

Manual data collection is still the gold standard for disease-specific patient registries. However, CAPRI-3 uses text mining (an artificial intelligence (AI) technology) for patient identification and data collection. The aim of this study is to demonstrate the reliability and efficiency of this AI-driven approach.

METHODS

CAPRI-3 is an observational retrospective multicenter cohort registry on metastatic prostate cancer. We tested the patient-identification algorithm and automated data extraction through manual validation of the same patients in two pilots in 2019 and 2022.

RESULTS

Pilot one identified 2030 patients and pilot two 9464 patients. The negative predictive value of the algorithm was maximized to prevent false exclusions and reached 94.8%. The completeness and accuracy of the automated data extraction were 92.3% or higher, except for date fields and inaccessible data (images/pdf) (10-88.9%). Additional manual quality control took over 3 h less time per patient than the original fully manual CAPRI registry (105 vs. 300 min).

CONCLUSIONS

The CAPRI-3 patient-identification algorithm is a sound replacement for excluding ineligible candidates. The AI-driven data extraction is largely accurate and complete, but manual quality control is needed for less reliable and inaccessible data. Overall, the AI-driven approach of the CAPRI-3 registry is reliable and timesaving.

Outcomes and prognosis of non-small cell lung cancer patients who underwent curable surgery: a protocol for a real-world, retrospective, population-based and nationwide Chinese National Lung Cancer Cohort (CNLCC) study

INTRODUCTION

Surgery is one of the main approaches for the comprehensive treatment of early and locally advanced non-small cell lung cancer (NSCLC). This study conducts a nationwide multicentre study to explore factors that could influence the outcomes of patients with I-IIIA NSCLC who underwent curable surgery in real-world scenarios.

METHODS AND ANALYSIS

All patients diagnosed with NSCLC between January 2013 and December 2020 will be identified from 30 large public medical services centres in mainland China. The algorithm of natural language processing and artificial intelligence techniques were used to extract data from electronic health records of enrolled patients who fulfil the inclusion criteria. Six categories of parameters are collected and stored from the electronic records, then the parameters will be structured as a high-quality structured case report form. The code book will be compiled and each parameter will be classified and designated a code. In addition, the study retrieves the survival status and causes of death of patients from the Chinese Centre for Disease Control and Prevention. The primary endpoints are overall survival and the secondary endpoint is disease-free survival. Finally, an online platform is formed for data queries and the original records will be stored as secure electronic documents.

ETHICS AND DISSEMINATION

The study has been approved by the Ethical Committee of the Chinese Academy of Medical Sciences. Study findings will be disseminated via presentations at conferences and publications in open-access journals. This study has been registered in the Chinese Trial Register (ChiCTR2100052773) on 11 May 2021, http://www.chictr.org.cn/showproj.aspx?proj=136659.

TRIAL REGISTRATION NUMBER

ChiCTR2100052773.

A text-mining approach to study the real-world effectiveness and potentially fatal immune-related adverse events of PD-1 and PD-L1 inhibitors in older patients with stage III/IV non-small cell lung cancer

BACKGROUND

This study was designed to investigate the impact of age on the effectiveness and immune-related adverse events (irAEs) of programmed death-(ligand)1 [PD-(L)1] inhibitors in patients with non-small cell lung cancer (NSCLC) using a novel text-mining technique.

METHODS

This retrospective study included patients with stage III/IV NSCLC treated with a PD-(L)1 inhibitor (nivolumab, pembrolizumab, atezolizumab and durvalumab) at Leiden University Medical Centre and Haga Teaching hospital, (both in The Netherlands) from September 2016 to May 2021. All the relevant data was extracted from the structured and unstructured fields of the Electronic Health Records using a novel text-mining tool. Effectiveness [progression-free survival (PFS) and overall survival (OS)] and safety (the incidence of nine potentially fatal irAEs and systemic corticosteroid requirement) outcomes were compared across age subgroups (young: < 65 years, Middle-aged: 65-74 years, and old: ≥ 75 years) after adjustment for confounding.

RESULTS

Of 689 patients, 310 patients (45.0%) were < 65 years, 275 patients (39.9%) were aged between 65 and 74 years, and 104 patients (15.1%) were ≥ 75 years. There was no significant difference between younger and older patients regarding PFS (median PFS 12, 8, 13 months respectively; Hazard ratio (HR)_middle-aged = 1.14, 95% CI 0.92-1.41; HR_old = 1.10, 95% CI 0.78-1.42). This was also the case for OS (median OS 19, 14, 18 months respectively; HR_middle-aged = 1.22, 95% CI 0.96-1.53; HR_old = 1.10, 95% CI 0.79-1.52). Safety analysis demonstrated a higher incidence of pneumonitis among patients aged 65-74. When all the investigated irAEs were pooled, there was no statistically significant difference found between age and the incidence of potentially fatal irAEs.

CONCLUSIONS

The use of PD-(L)1 inhibitors is not associated with age related decrease of PFS and OS, nor with increased incidence of serious irAEs compared to younger patients receiving these treatments. Chronological age must therefore not be used as a predictor for the effectiveness or safety of ICIs.

External Validation of Prediction Models for Surgical Complications in People Considering Total Hip or Knee Arthroplasty Was Successful for Delirium but Not for Surgical Site Infection, Postoperative Bleeding, and Nerve Damage: A Retrospective Cohort Study

Although several models for the prediction of surgical complications after primary total hip or total knee replacement (THA and TKA, respectively) are available, only a few models have been externally validated. The aim of this study was to externally validate four previously developed models for the prediction of surgical complications in people considering primary THA or TKA. We included 2614 patients who underwent primary THA or TKA in secondary care between 2017 and 2020. Individual predicted probabilities of the risk for surgical complication per outcome (i.e., surgical site infection, postoperative bleeding, delirium, and nerve damage) were calculated for each model. The discriminative performance of patients with and without the outcome was assessed with the area under the receiver operating characteristic curve (AUC), and predictive performance was assessed with calibration plots. The predicted risk for all models varied between <0.01 and 33.5%. Good discriminative performance was found for the model for delirium with an AUC of 84% (95% CI of 0.82-0.87). For all other outcomes, poor discriminative performance was found; 55% (95% CI of 0.52-0.58) for the model for surgical site infection, 61% (95% CI of 0.59-0.64) for the model for postoperative bleeding, and 57% (95% CI of 0.53-0.61) for the model for nerve damage. Calibration of the model for delirium was moderate, resulting in an underestimation of the actual probability between 2 and 6%, and exceeding 8%. Calibration of all other models was poor. Our external validation of four internally validated prediction models for surgical complications after THA and TKA demonstrated a lack of predictive accuracy when applied in another Dutch hospital population, with the exception of the model for delirium. This model included age, the presence of a heart disease, and the presence of a disease of the central nervous system as predictor variables. We recommend that clinicians use this simple and straightforward delirium model during preoperative counselling, shared decision-making, and early delirium precautionary interventions.

Application of Electronic Health Record Text Mining: Real-World Tolerability, Safety, and Efficacy of Adjuvant Melanoma Treatments

Introduction: Nivolumab (N), pembrolizumab (P), and dabrafenib plus trametinib (D + T) have been registered as adjuvant treatments for resected stage III and IV melanoma since 2018. Electronic health records (EHRs) are a real-world data source that can be used to review treatments in clinical practice. In this study, we applied EHR text-mining software to evaluate the real-world tolerability, safety, and efficacy of adjuvant melanoma treatments. Methods: Adult melanoma patients receiving adjuvant treatment between January 2019 and October 2021 at the Leiden University Medical Center, the Netherlands, were included. CTcue text-mining software (v3.1.0, CTcue B.V., Amsterdam, The Netherlands) was used to construct rule-based queries and perform context analysis for patient inclusion and data collection from structured and unstructured EHR data. Results: In total, 122 patients were included: 54 patients treated with nivolumab, 48 with pembrolizumab, and 20 with D + T. Significantly more patients discontinued treatment due to toxicity on D + T (N: 16%, P: 6%, D + T: 40%), and X2 (6, n = 122) = 14.6 and p = 0.024. Immune checkpoint inhibitors (ICIs) mainly showed immune-related treatment-limiting adverse events (AEs), and chronic thyroid-related AE occurred frequently (hyperthyroidism: N: 15%, P: 13%, hypothyroidism: N: 20%, P: 19%). Treatment-limiting toxicity from D + T was primarily a combination of reversible AEs, including pyrexia and fatigue. The 1-year recurrence-free survival was 70.3% after nivolumab, 72.4% after pembrolizumab, and 83.0% after D + T. Conclusions: Text-mining EHR is a valuable method to collect real-world data to evaluate adjuvant melanoma treatments. ICIs were better tolerated than D + T, in line with RCT results. For BRAF+ patients, physicians must weigh the higher risk of reversible treatment-limiting AEs of D + T against the risk of long-term immune-related AEs.

Artificial intelligence in pharmacology research and practice

In recent years, the use of artificial intelligence (AI) in health care has risen steadily, including a wide range of applications in the field of pharmacology. AI is now used throughout the entire continuum of pharmacology research and clinical practice and from early drug discovery to real-world datamining. The types of AI models used range from unsupervised clustering of drugs or patients aimed at identifying potential drug compounds or suitable patient populations, to supervised machine learning approaches to improve therapeutic drug monitoring. Additionally, natural language processing is increasingly used to mine electronic health records to obtain real-world data. In this mini-review, we discuss the basics of AI followed by an outline of its application in pharmacology research and clinical practice.

Real-World Metastatic Renal Cell Carcinoma Treatment Patterns and Clinical Outcomes in The Netherlands

The number of treatment options for patients with metastatic renal cell carcinoma (mRCC) has significantly grown in the last 15 years. Although randomized controlled trials are fundamental in investigating mRCC treatment efficacy, their external validity can be limited. Therefore, the efficacy of the different treatment options should also be evaluated in clinical practice. We performed a chart review of electronic health records using text mining software to study the current treatment patterns and outcomes. mRCC patients from two large hospitals in the Netherlands, starting treatment between January 2015 and May 2020, were included. Data were collected from electronic health records using a validated text mining tool. Primary endpoints were progression-free survival (PFS) and overall survival (OS). Statistical analyses were performed using the Kaplan-Meier method. Most frequent first-line treatments were pazopanib (n = 70), sunitinib (n = 34), and nivolumab with ipilimumab (n = 28). The overall median PFS values for first-line treatment were 15.7 months (95% confidence interval [95%CI], 8.8-20.7), 16.3 months (95%CI, 9.3-not estimable [NE]) for pazopanib, and 6.9 months (95% CI, 4.4-NE) for sunitinib. The overall median OS values were 33.4 months (95%CI, 28.1-50.9 months), 39.3 months (95%CI, 29.5-NE) for pazopanib, and 28.1 months (95%CI, 7.0-NE) for sunitinib. For nivolumab with ipilimumab, median PFS and median OS were not reached. Of the patients who finished first- and second-line treatments, 64 and 62% received follow-up treatments, respectively. With most patients starting on pazopanib and sunitinib, these real-world treatment outcomes were most likely better than in pivotal trials, which may be due to extensive follow-up treatments.

Real-world evaluation of supportive care using an electronic health record text-mining tool: G-CSF use in breast cancer patients

PURPOSE

Chemotherapy-induced febrile neutropenia (FN) is a life-threatening and chemotherapy dose-limiting adverse event. FN can be prevented with granulocyte-colony stimulating factors (G-CSFs). Guidelines recommend primary G-CSF use for patients receiving either high (> 20%) FN risk (HR) chemotherapy, or intermediate (10-20%) FN risk (IR) chemotherapy if the overall risk with additional patient-related risk factors exceeds 20%. In this study, we applied an EHR text-mining tool for real-world G-CSF treatment evaluation in breast cancer patients.

METHODS

Breast cancer patients receiving IR or HR chemotherapy treatments between January 2015 and February 2021 at LUMC, the Netherlands, were included. We retrospectively collected data from EHR with a text-mining tool and assessed G-CSF use, risk factors, and the FN and neutropenia (grades 3-4) and incidence.

RESULTS

A total of 190 female patients were included, who received 77 HR and 113 IR treatments. In 88.3% of the HR regimens, G-CSF was administered; 7.3% of these patients developed FN vs. 33.3% without G-CSF. Although most IR regimen patients had ≥ 2 risk factors, only 4% received G-CSF, of which none developed neutropenia. However, without G-CSF, 11.9% developed FN and 31.2% severe neutropenia.

CONCLUSIONS

Our text-mining study shows high G-CSF use among HR regimen patients, and low use among IR regimen patients, although most had ≥ 2 risk factors. Therefore, current practice is not completely in accordance with the guidelines. This shows the need for increased awareness and clarity regarding risk factors. Also, text-mining can effectively be implemented for the evaluation of patient care.

Treosulfan-induced myalgia in pediatric hematopoietic stem cell transplantation identified by an electronic health record text mining tool

Treosulfan is increasingly used as myeloablative agent in conditioning regimen prior to allogeneic hematopoietic stem cell transplantation (HSCT). In our pediatric HSCT program, myalgia was regularly observed after treosulfan-based conditioning, which is a relatively unknown side effect. Using a natural language processing and text-mining tool (CDC), we investigated whether treosulfan compared with busulfan was associated with an increased risk of myalgia. Furthermore, among treosulfan users, we studied the characteristics of given treatment of myalgia, and studied prognostic factors for developing myalgia during treosulfan use. Electronic Health Records (EHRs) until 28 days after HSCT were screened using the CDC for myalgia and 22 synonyms. Time to myalgia, location of pain, duration, severity and drug treatment were collected. Pain severity was classified according to the WHO pain relief ladder. Logistic regression was performed to assess prognostic factors. 114 patients received treosulfan and 92 busulfan. Myalgia was reported in 37 patients; 34 patients in the treosulfan group and 3 patients in the busulfan group (p = 0.01). In the treosulfan group, median time to myalgia was 7 days (0–12) and median duration of pain was 19 days (4–73). 44% of patients needed strong acting opiates and adjuvant medicines (e.g. ketamine). Hemoglobinopathy was a significant risk factor, as compared to other underlying diseases (OR 7.16 95% CI 2.09–30.03, p = 0.003). Myalgia appears to be a common adverse effect of treosulfan in pediatric HSCT, especially in hemoglobinopathy. Using the CDC, EHRs were easily screened to detect this previously unknown side effect, proving the effectiveness of the tool. Recognition of treosulfan-induced myalgia is important for adequate pain management strategies and thereby for improving the quality of hospital stay.

Discontinuation of Antihypertensive Medications on the Outcome of Hospitalized Patients With Severe Acute Respiratory Syndrome-Coronavirus 2

Supplemental Digital Content is available in the text.

RAASi (renin-angiotensin-aldosterone system inhibitors) are suggested as possible treatment option in the early phase of severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2) infection. A meta-analysis investigating the possible detrimental effects of RAASi on the severity of (SARS-CoV-2) infection showed that ambulatory use of RAASi, by hospitalized patients, has a neutral effect. It is, however, conceivable that this observation is biased by the fact that antihypertensive medications, are often discontinued at or during admission in hospitalized patients with SARS-CoV-2. We, therefore, investigated the effect of discontinuation of antihypertensive medications, in hospitalized patients with SARS-CoV-2. We performed a retrospective observational study on 1584 hospitalized patients with SARS-CoV-2 from 10 participating hospitals in the Netherlands. Differences in the outcome (severity of disease or death) between the groups in which medications were either continued or discontinued during the course of hospitalization were assessed using logistic regression models. Discontinuation of angiotensin receptor blockers, ACE (angiotensin-converting enzyme) inhibitors and β-blockers, even when corrected for sex, age, and severity of symptoms during admission, resulted in a 2 to 4× higher risk of dying from SARS-CoV-2 infection (odds ratio [95% CI]); angiotensin receptor blockers 2.65 [1.17–6.04], ACE inhibitor (2.28 [1.15–4.54]), and β-blocker (3.60 [1.10–10.27]). In conclusion, discontinuation of at-home ACE inhibitor, angiotensin receptor blockers, or β-blocker in patients hospitalized for a SARS-CoV-2 infection was associated with an increased risk of dying, whereas discontinuation of calcium channel blockers and diuretics was not.