A 52-Year-Old Woman with Abdominal Pain and Vomiting.

A 52-Year-Old Woman with Abdominal Pain and VomitingA 52-year-old woman presented for evaluation of abdominal pain, nausea, and vomiting after consuming a large calzone. How do you approach the evaluation, and what is the diagnosis?

Translation and Psychometric Evaluation of the Hindi Language Version of the Pediatric Nausea Assessment Tool (PeNAT) in the Indian Population.

OBJECTIVES: To translate the Pediatric Nausea Assessment Tool (PeNAT) into Hindi and validate it in Indian pediatric cancer patients and survivors. METHODS: The PeNAT-Hindi was finalized by forward and backward translations, and pilot testing. The PeNAT-Hindi was administered to 200 Hindi-speaking pediatric (4-18 y) cancer patients/survivors, in three groups. These included pediatric cancer patients who had recently received chemotherapy (n = 150); who received no chemotherapy within 5 d (n = 25) and survivors (n = 25). Construct validity was tested by comparing scores among the three groups. Test-retest reliability and criterion validity were estimated by the correlation of the first PeNAT score with the second (taken 1 h later) PeNAT score and the number of vomiting/retching episodes, respectively. Convergent validity and discriminant validity were estimated by correlating PeNAT scores with parent-assessed nausea severity, and pain, respectively. The responsiveness was tested by comparing second PeNAT scores with subsequent divergent PeNAT scores among patients reporting subjective change (improvement and worsening, respectively) in nausea severity. RESULTS: Test-retest reliability of PeNAT-Hindi was good (intraclass correlation = 0.791). The initial PeNAT score had moderate correlation with the number of vomiting/retching episodes (Spearman ρ = 0.401). Median PeNAT scores in group 1 versus groups 2 and 3 were significantly different (p < 0.001). Initial PeNAT scores showed a moderate correlation with parent-assessed nausea (Spearman ρ = 0.657) and a weak correlation with parent-assessed pain (Spearman ρ = 0.319). The responsiveness (standardized response mean) of PeNAT-Hindi to the change in nausea severity was -1.79 (improvement) and 2.19 (worsening), respectively. CONCLUSION: PeNAT-Hindi showed good reliability and acceptable validity. It may be used among Hindi-speaking children for measuring nausea. The responsiveness of PeNAT-Hindi needs further evaluation.

A deep learning-based interpretable decision tool for predicting high risk of chemotherapy-induced nausea and vomiting in cancer patients prescribed highly emetogenic chemotherapy.

OBJECTIVE: This study aims to develop a risk prediction model for chemotherapy-induced nausea and vomiting (CINV) in cancer patients receiving highly emetogenic chemotherapy (HEC) and identify the variables that have the most significant impact on prediction. METHODS: Data from Tianjin Medical University General Hospital were collected and subjected to stepwise data preprocessing. Deep learning algorithms, including deep forest, and typical machine learning algorithms such as support vector machine (SVM), categorical boosting (CatBoost), random forest, decision tree, and neural network were used to develop the prediction model. After training the model and conducting hyperparameter optimization (HPO) through cross-validation in the training set, the performance was evaluated using the test set. Shapley additive explanations (SHAP), partial dependence plot (PDP), and Local Interpretable Model-Agnostic Explanations (LIME) techniques were employed to explain the optimal model. Model performance was assessed using AUC, F1 score, accuracy, specificity, sensitivity, and Brier score. RESULTS: The deep forest model exhibited good discrimination, outperforming typical machine learning models, with an AUC of 0.850 (95%CI, 0.780-0.919), an F1 score of 0.757, an accuracy of 0.852, a specificity of 0.863, a sensitivity of 0.784, and a Brier score of 0.082. The top five important features in the model were creatinine clearance (Ccr), age, gender, anticipatory nausea and vomiting, and antiemetic regimen. Among these, Ccr had the most significant predictive value. The risk of CINV decreased with increased Ccr and age, while it was higher in the presence of anticipatory nausea and vomiting, female gender, and non-standard antiemetic regimen. CONCLUSION: The deep forest model demonstrated good discrimination in predicting the risk of CINV in cancer patients prescribed HEC. Kidney function, as represented by Ccr, played a crucial role in the model's prediction. The clinical application of this predictive tool can help assess individual risks and improve patient care by proactively optimizing the use of antiemetics in cancer patients receiving HEC.

Nausea is the only symptom associated with delayed gastric emptying in pediatric functional dyspepsia patients.

OBJECTIVES: Functional dyspepsia (FD) is a heterogeneous functional gastrointestinal disorder (FGID) with a highly prevalent symptom complex. The aim of our study is to investigate the relation between symptoms of FD and results of gastric emptying (GE) breath test in children. METHODS: This study included patients (6-17 years old) presented at the general gastroenterology outpatient clinic with dyspeptic symptoms (Rome IV criteria) and underwent careful history taking with clinical examination. A GE breath test with a 13 C-octanoic acid labeled (250 kcal) solid meal was performed and dyspepsia symptom scores, clarified using pictograms for postprandial fullness, bloating, belching, nausea, vomiting, epigastric pain and burning, were obtained every 15 min ranging from 0 to 4 for a total of 240 min. The severity of the complaints (overall and individual symptoms) as displayed by the symptom questionnaire was compared between normal and delayed GE groups. The relationship between GE time and the severity of FD symptoms was assessed using Mann-Whitney test. RESULTS: Thirty nine FD patients (55% girls; mean age: 11.9 ± 3.3 years) participated in the study. Of these, 43% had delayed GE. The overall symptom severity in patients with delayed GE was similar to the symptoms of patients with a normal GE rate (149.5 ± 12.7 points vs. 123.9 ± 9.0; p = 0.19). Individual symptoms scores showed only nausea to be significantly increased in the group with delayed GE (21.5 ± 1.9 points vs. 33.2 ± 4.6; p = 0.048, p < 0.1). CONCLUSION: Especially in children with nausea as presenting symptom of FD, a low threshold should be withheld to perform a GE breath test.

[Analysis of the clinical characteristics and misdiagnosis of area postrema syndrome manifesting as intractable nausea, vomiting, and hiccups in neuromyelitis optica spectrum disorders].

Objective: To investigate the misdiagnosis of area postrema syndrome (APS) manifesting as intractable nausea, vomiting and hiccups in neuromyelitis optic spectrum disease (NMOSD) and reduce the risk of misdiagnosis. Methods: We retrospectively analyzed data from NMOSD patients attending the Department of Neurology at the First Medical Center of PLA General Hospital between January 2019 and July 2021. SPSS25.0 was then used to analyze the manifestations, misdiagnosis, and mistreatment of APS. Results: A total of 207 patients with NMOSD were included, including 21 males and 186 females. The mean age of onset was 39±15 years (range: 5-72 years). The proportion of patients who were positive for serum aquaporin 4 antibody was 82.6% (171/207). In total, 35.7% (74/207) of the NMOSD patients experienced APS during the disease course; of these patients, 70.3% (52/74) had APS as the first symptom and 29.7% (22/74) had APS as a secondary symptom. The misdiagnosis rates for these conditions were 90.4% (47/52) and 50.0% (11/22), respectively. As the first symptom, 19.2% (10/52) of patients during APS presented only with intractable nausea, vomiting and hiccups; 80.8% (42/52) of patients experienced other neurological symptoms. The Departments of Gastroenterology and General Medicine were the departments that most frequently made the first diagnosis of APS, accounting for 54.1% and 17.6% of patients, respectively. The most common misdiagnoses related to diseases of the digestive system and the median duration of misdiagnosis was 37 days. Conclusions: APS is a common symptom of NMOSD and is associated with a high rate of misdiagnosis. Other concomitant symptoms often occur with APS. Gaining an increased awareness of this disease/syndrome, obtaining a detailed patient history, and performing physical examinations are essential if we are to reduce and avoid misdiagnosis.

Waterpipe Tobacco (Hookah) Use in Pregnancy: Associations with Nausea and Vomiting of Pregnancy.

OBJECTIVE: Waterpipe tobacco (WPT) use is common among reproductive age patients and is often perceived as safer than cigarette use. Prior studies have shown a decrease in nausea and vomiting symptoms among pregnant women who use cigarettes, but no studies to date have examined these symptoms in pregnant women who use WPT. This study was aimed to investigate the extent of symptoms of nausea/vomiting of pregnancy among participants who self-reported WPT use during pregnancy. STUDY DESIGN: Secondary analysis of a prospective cohort study examining WPT use during pregnancy. Participants completed the Pregnancy-Unique Quantification of Emesis (PUQE) during first and third trimesters. Medical conditions were determined by medical record review. Participants were evaluated by sole WPT use versus dual/polysubstance WPT use and frequency of WPT use. RESULTS: Ninety-nine (100%) participants completed the PUQE questionnaire during first trimester and 82 (82.8%) completed the PUQE during third trimester. Almost all (91.9%) participants reported moderate nausea/vomiting symptoms at both assessments. There was no difference in frequency of WPT use in pregnancy or rates of dual/polysubstance WPT use in participants with all levels of the PUQE questionnaire. There was also no difference in rates of WPT use or PUQE scores between sole WPT users and dual/polysubstance users. When comparing low and high WPT use, those who were in the higher frequency use group had higher waterpipe dependence scale scores (7.2 vs. 5.3, p < 0.02). With regard to maternal medical comorbidities, the only difference between groups was that sole WPT users were more likely to have a diagnosis of asthma than dual/polysubstance users (36.8 vs. 14.9%, p < 0.02). CONCLUSION: There were no differences in symptoms of nausea and vomiting of pregnancy or medical conditions in pregnant women who use WPT with any frequency during pregnancy. However, sole WPT users had higher rates of asthma than dual/polysubstance WPT users. KEY POINTS: · Waterpipe tobacco use is one of the most common forms of tobacco use among reproductive age patients.. · Waterpipe tobacco use was not associated with any changes in nausea/vomiting of pregnancy symptoms.. · Future research on the use of waterpipe tobacco in pregnancy can aid in public health responses..

Diagnosis and Management of Cyclic Vomiting Syndrome: A Critical Review.

Cyclic vomiting syndrome (CVS) is a chronic disorder of gut-brain interaction characterized by recurrent disabling episodes of nausea, vomiting, and abdominal pain. CVS affects both children and adults with a prevalence of approximately 2% in the United States. CVS is more common in female individuals and affects all races. The pathophysiology of CVS is unknown and a combination of genetic, environmental, autonomic, and neurohormonal factors is believed to play a role. CVS is also closely associated with migraine headaches and likely have a shared pathophysiology. The diagnosis of CVS is based on the Rome criteria, and minimal recommended testing includes an upper endoscopy and imaging studies of the abdomen. CVS is frequently associated with anxiety, depression, and autonomic dysfunction. Patients with CVS commonly use cannabis therapeutically for symptom relief. By contrast, cannabinoid hyperemesis syndrome is believed to be a subset of CVS with chronic heavy cannabis use leading to hyperemesis. Due to the recalcitrant nature of the illness, patients often visit the emergency department and are hospitalized for acute CVS flares. Guidelines on the management of CVS recommend a biopsychosocial approach. Prophylactic therapy consists of tricyclic antidepressants (amitriptyline), antiepileptics (topiramate), and aprepitant in refractory patients. Abortive therapy consists of triptans, antiemetics (ondansetron), and sedation. Treatment of comorbid conditions is extremely important to improve overall patient outcomes. CVS has a significant negative impact on patients, families, and the healthcare system, and future research to understand its pathophysiology and develop targeted therapies is needed.

Recalling the severity of nausea and vomiting of pregnancy - a study using Pregnancy-Unique Quantification of Emesis Questionnaire.

The accuracy of the recall of the severity of nausea and vomiting of pregnancy (NVP) with Pregnancy-Unique Quantification of Emesis (PUQE) questionnaire has been questioned. We aimed to compare PUQE scores of women recalling the worst episode of NVP of their current pregnancy in different gestational weeks (gwks). Total of 2343 pregnant women (gwks 7-40) were recruited. Four groups were formed according to the gwks at reply: ≤16 gwks (n = 554), ≤20 gwks (n = 1209), >20 gwks (n = 1134) and ≥24 gwks (n = 495). PUQE scores were similar between the groups. Consequently, consistency of PUQE scores across the groups endorses the useability of the PUQE questionnaire in retrospective assessment of the overall severity of NVP in different gwks, regardless of passing of the peak NVP symptoms.Impact statementWhat is already known on this subject? Retrospective evaluation of the severity of nausea and vomiting of pregnancy (NVP) has been argued to be disposed to recall bias. Structured Pregnancy-Unique Quantification of Emesis (PUQE) questionnaire is a validated tool for assessing the severity of NVP.What do the results of this study add? When the women recalled the most severe NVP symptoms of their current pregnancy, no differences in the PUQE scores were found despite different gestational weeks at reply. Of distinct PUQE questions, women answering in early pregnancy reported longer duration of nausea than women answering in late pregnancy, but other questions were rated similarly.What the implications are of these findings for clinical practice and/or further research? Our aim was to compare the PUQE scores between the women who filled in the PUQE questionnaire in early or in late pregnancy, instructed to recall their worst symptoms in their current pregnancy. As there were no differences between the groups in total PUQE scores, our results support the application of PUQE questionnaire to assess the severity of NVP during pregnancy not only concurrent to the peak symptoms but also retrospectively.

Recurrent Nausea and Vomiting with Weight Loss Associated with Hypothyroidism: Fact or Myth.

BACKGROUND: Hypothyroidism is a commonly encountered endocrine disorder presenting in various clinical settings. It usually presents with classic manifestations, which are readily recognized and, therefore, easy to diagnose. However, occasionally, patients present with unusual symptoms, which becomes a challenge to diagnose. Thyroid dysfunction affects many body organs, including the gut and viscera. Studies show that intestinal motility might be affected by multiple factors, such as neuromuscular dysfunction, myopathy, or alterations in hormone receptors. CASE PRESENTATION: Here, we present the first case of a 21-year-old female student who had complaints of recurrent nausea, vomiting, loose stool, abdominal pain, and weight loss. In the second case, a 25-year-old male student presented with recurrent nausea, vomiting, loose stool, and weight loss. Their unremarkable blood routines and gastrointestinal-specific investigations failed to ascertain the diagnosis. Later, primary hypothyroidism was established by typical biochemical abnormalities. CONCLUSION: Thyroxine replacement treatment successfully resolved the presenting symptoms and normalized biochemical reports.

Multichannel magnetogastrogram: a clinical marker for pediatric chronic nausea.

Chronic nausea is a widespread functional disease in children with numerous comorbidities. High-resolution electrogastrogram (HR-EGG) has shown sufficient sensitivity as a noninvasive clinical marker to objectively detect distinct gastric slow wave properties in children with functional nausea. We hypothesized that the increased precision of magnetogastrogram (MGG) slow wave recordings could provide supplementary information not evident on HR-EGG. We evaluated simultaneous pre- and postprandial MGG and HR-EGG recordings in pediatric patients with chronic nausea and healthy asymptomatic subjects, while also measuring nausea intensity and nausea severity. We found significant reductions in postprandial dominant frequency and normogastric power, and higher levels of postprandial bradygastric power in patients with nausea in both MGG and HR-EGG. MGG also detected significantly lower preprandial normogastric power in patients. A significant difference in the mean preprandial gastric slow wave propagation direction was observed in patients as compared with controls in both MGG (control: 180 ± 61°, patient: 34 ±72°; P < 0.05) and HR-EGG (control: 240 ± 39°, patient: 180 ± 46°; P < 0.05). Patients also showed a significant change in the mean slow wave direction between pre- and postprandial periods in MGG (P < 0.05). No statistical differences were observed in propagation speed between healthy subjects and patients in either MGG or HR-EGG pre/postprandial periods. The use of MGG and/or HR-EGG represents an opportunity to assess noninvasively the effects of chronic nausea on gastric slow wave activity. MGG data may offer the opportunity for further refinement of the more portable and economical HR-EGG in future machine-learning approaches for functional nausea.NEW & NOTEWORTHY Pediatric chronic nausea is a difficult-to-measure subjective complaint that requires objective diagnosis, clinical assessment, and individualized treatment plans. Our study demonstrates that multichannel MGG used in conjunction with custom HR-EGG detects key pathological signatures of functional nausea in children. This quantifiable measure may allow more personalized diagnosis and treatment in addition to minimizing the cost and potential radiation associated with current diagnostic approaches.

Worldwide prevalence and burden of gastroparesis-like symptoms as defined by the United European Gastroenterology (UEG) and European Society for Neurogastroenterology and Motility (ESNM) consensus on gastroparesis.

BACKGROUND/OBJECTIVES: The global epidemiology of gastroparesis is unknown. The European UEG and European Society for Neurogastroenterology and motility consensus defines Gastroparesis as a condition characterized by delayed gastric emptying in the absence of mechanical obstruction, with a symptom pattern of nausea and/or vomiting and overlapping postprandial distress syndrome (PDS). Real-world evidence of this gastroparesis-like symptom pattern is a crucial step in understanding the epidemiology of gastroparesis. METHODS: In the Rome Foundation Global Epidemiology Study, 54,127 respondents from 26 countries completed the Rome IV Diagnostic Questionnaire and variables associated with disorders of gut-brain interaction via Internet. We selected subjects with gastroparesis-like symptoms (GPLS) (nausea and/or vomiting ≥1 day/week and simultaneous PDS). Patients reporting organic gastrointestinal disease, or fulfilling criteria for self-induced vomiting, cyclic vomiting or cannabinoid hyperemesis syndrome were excluded. We determined prevalence, associated comorbidities, quality of life (QoL) (PROMIS Global-10), symptoms of anxiety and depression (PHQ-4), somatic symptoms (PHQ-12), and healthcare utilization. RESULTS: The global prevalence of GPLS was 0.9% overall and 1.3% among diabetic individuals. Subjects with GPLS showed frequent overlapping of epigastric pain syndrome and irritable bowel syndrome. Subjects with GPLS had significantly lower body mass index, QoL, more non-gastrointestinal somatic complaints, symptoms of anxiety and depression, higher medication usage and doctor visits in the overall and diabetic population, compared to subjects without these symptoms. CONCLUSIONS: GPLS are common worldwide and more common in diabetic patients. The symptom complex is associated with multiple aspects of illness and an increased healthcare consumption.

Evaluation of a nausea assessment tool in the pediatric oncology population.

PURPOSE: Nausea is a symptom that is often experienced but misunderstood. Its impact is amplified in pediatric oncology patients. Nausea assessments in pediatric oncology are few and not yet widely used. The Baxter Retching Faces (BARF) scale holds promise and is used in some pediatric oncology units. The purpose of this evidence-based practice project was to evaluate the impact of the BARF scale on nursing assessments in inpatient pediatric hematology/oncology and stem cell transplant settings. CONCLUSIONS: Project work took place on one hematology/oncology unit and one stem cell transplant unit. Thirty nurses completed the acceptability of intervention measure (AIM), intervention appropriateness measure (IAM), and feasibility of intervention measure (FIM) to provide initial feedback on the BARF scale. Nurses used the BARF scale over an 8-week period. Patient demographics and BARF scores were collected during nausea assessments. Twenty nurses completed AIM/IAM/FIM scores postintervention. There were no significant changes in scores, though nurses stated that the tool was easy to use and provided valuable symptom feedback. PRACTICE IMPLICATIONS: Symptom management will continue to be a challenge in the pediatric setting given developmental and disease-specific considerations inherent to the specialty. While this small-scale project did not achieve statistical significance, the evidence and feedback from nursing staff present a compelling case that efforts to investigate and integrate improved methods of assessing nausea and other problematic symptoms are needed to enhance nursing practice and impact patient-centered outcomes.

Use of Cluster Analysis to Identify Subtypes of Pediatric Functional Nausea.

OBJECTIVE: The aim of the study was to evaluate whether there are clinical subtypes in children with functional nausea based on comorbidities and responses to the Nausea Profile questionnaire. METHODS: Patients from the Neurointestinal and Motility Program clinical registry at Lurie Children's Hospital were included if they met Rome IV criteria for functional nausea. Patients completed the Nausea Profile, a multidimensional measure of nausea with gastrointestinal, emotional, and somatic subscales. Comorbidities were assessed by chart review and self-report measures. Latent class analysis was used to identify patient groups based on comorbidities. To assess if model-identified groups were predictive of differences in nausea quality, Nausea Profile subscale means were compared between groups and used to predict group membership. Conversely, k-means analysis was used to divide the sample into groups based upon Nausea Profile subscale scores, to determine if identified groups had different comorbidities. RESULTS: Seventy-two patients (n = 53 girls) with a mean age (±SD) 14.5 ±â€Š2.9 were included. Two clinical subtypes were identified based on comorbidities, with responses on the emotional subscale of the Nausea Profile predicting group membership (P < 0.04). When patients were grouped by nausea quality, the resulting clusters differed on psychiatric comorbidities (P < 0.001). CONCLUSIONS: Our findings support the existence of nausea subtypes within the broad diagnosis of functional nausea. One such subtype is an emotional predominant nausea supporting the notion that anxiety and depression constitute a subset of patients with nausea. Thus, patients may benefit from a treatment approach that integrates both GI assessment and psychiatric support in their care.

Psychometric properties of the Pregnancy-Unique Quantification of Emesis (PUQE-24) Scale.

The objective of this study is to conduct a reliability and validity study of the Turkish version of the 'Pregnancy-Unique Quantification of Emesis (PUQE-24)' in pregnant women. In the Turkish version, Cronbach's alpha coefficient was 0.75, and the item-total score correlations were between 0.75 and 0.85. In the exploratory factor analysis it was determined that the scale had a single-factor structure explaining 65.968% of the total variance. The factor load values of the scale were found to be between 0.776 and 0.831. The Turkish version of scale was found to be a valid and reliable measurement in pregnant women.IMPACT STATEMENTWhat is already known on this subject? Complaints of nausea and vomiting during pregnancy are common. Therefore, it is important to evaluate nausea and vomiting during pregnancy with a valid and reliable tool. Pregnancy-Unique Quantification of Emesis (PUQE-24) is a reliable tool for assessing the severity of nausea and vomiting symptoms. The original scale was translated into various languages, its validity and reliability were made in some countries and it was used in many studies.What do the results of this study add? The present study showed that Pregnancy-Unique Quantification of Emesis (PUQE-24) is valid and reliable for Turkish pregnant women. Thus, the scale can be used as a reliable tool in Turkish population.What are the implications of these findings for clinical practice and/or further research? As a result of this study, Pregnancy-Unique Quantification of Emesis (PUQE-24) can be used as a validated tool for the Turkish population during clinical practice by healthcare professionals and researchers, who are evaluating nausea and vomiting during pregnancy. In future studies, it can be used as an objective assessment tool to determine whether an intervention is needed for nausea and vomiting during pregnancy or to reveal whether the intervention has worked.

A Practical 5-Step Approach to Nausea and Vomiting.

Nausea and vomiting (N/V) are common presenting complaints in the outpatient and inpatient settings. These symptoms can be associated with high morbidity and poor quality of life, particularly in those with chronic symptoms. The clinical approach to N/V can be challenging, given the numerous possible underlying causes as well as the vast array of diagnostic and therapeutic options. In this concise review, we provide a practical 5-step approach to the clinical evaluation and treatment of N/V, suitable for application in the primary care and subspecialty settings. The 5-step approach includes (1) defining what the patient means by N/V, (2) determining whether symptoms are acute or chronic, (3) considering medication or toxin adverse effects, (4) using the patient's presentation, severity of symptoms, and physical examination findings to formulate a differential diagnosis and to guide evaluation, and (5) directing treatment on the basis of knowledge of neurotransmitters and receptors involved in the emetic pathways. We discuss the pathophysiology (neuronal pathways and neurotransmitters), differential diagnosis (medication and toxin adverse effects, neurologic causes, gastrointestinal diseases, metabolic and endocrine conditions, and psychogenic disorders), initial evaluation and risk stratification, and management and treatment options. Management of symptoms that are acute in onset or mild in severity may involve an empirical trial of antiemetics without extensive testing. In contrast, when symptoms are chronic or moderate-severe, testing for an underlying cause should be performed, and medication adverse effects, neurologic causes, gastrointestinal diseases, metabolic or endocrine conditions, and psychogenic disorders should be considered in particular.

Developing the Nausea and Vomiting Thermometer Scale in children with cancer.

BACKGROUND: This study aimed to develop the Nausea and Vomiting Thermometer Scale (NVTS) in children with cancer. METHODS: This methodological study was conducted on 250 children with cancer at the research and training university hospital in Turkey between September 2019 and January 2020. The t-test, the ROC analysis, the Diagnostic index, and the Youden index were used for determining the scale of the cutting point. The regression analysis, the intra-class correlation coefficient, and the BlandAltman analysis were used for the data analysis. RESULTS: The scale-level content validity index was .94, which was coherent. As a result of the ROC analysis, the cut-off point was determined as three points. The NVTS showed good reliability, with an intra-class correlation coefficient of .99. In the linear regression analysis, a model was created for chemotherapy drugs, nausea and vomiting type, vomiting status, and the number of children with cancer who vomited explained 44.9% of their nausea and vomiting status. The results of the Bland-Altman analysis showed that the correlation coefficient between the differences and the means was insignificant.

Identification of Uncontrolled Symptoms in Cancer Patients Using Natural Language Processing.

CONTEXT: For patients with cancer, uncontrolled pain and other symptoms are the leading cause of unplanned hospitalizations. Early access to specialty palliative care (PC) is effective to reduce symptom burden, but more efficient approaches are needed for rapid identification and referral. Information on symptom burden largely exists in free-text notes, limiting its utility as a trigger for best practice alerts or automated referrals. OBJECTIVES: To evaluate whether natural language processing (NLP) can be used to identify uncontrolled symptoms (pain, dyspnea, or nausea/vomiting) in the electronic health record (EHR) among hospitalized cancer patients with advanced disease. METHODS: The dataset included 1,644 hospitalization encounters for cancer patients admitted from 1/2017 -6/2019. We randomly sampled 296 encounters, which included 15,580 clinical notes. We manually reviewed the notes and recorded symptom severity. The primary endpoint was an indicator for whether a symptom was labeled as "controlled" (none, mild, not reported) or as "uncontrolled" (moderate or severe). We randomly split the data into training and test sets and used the Random Forest algorithm to evaluate final model performance. RESULTS: Our models predicted presence of an uncontrolled symptom with the following performance: pain with 61% accuracy, 69% sensitivity, and 46% specificity (F1: 69.5); nausea/vomiting with 68% accuracy, 21% sensitivity, and 90% specificity (F1: 29.4); and dyspnea with 80% accuracy, 22% sensitivity, and 88% specificity (F1: 21.1). CONCLUSION: This study demonstrated initial feasibility of using NLP to identify hospitalized cancer patients with uncontrolled symptoms. Further model development is needed before these algorithms could be implemented to trigger early access to PC.