Can we accurately identify chemotherapy-related acute care visits in administrative data?

185 Background: Administrative data is increasingly being used to study treatment related complications that lead to acute care visits such as emergency department visits or hospitalizations (ED+H). We evaluated the accuracy of diagnosis codes for identifying chemotherapy related acute care visits (CRVs) among women with breast cancer. Methods: We prospectively developed algorithms to identify CRVs from administrative data in women receiving adjuvant chemotherapy for breast cancer in Ontario, Canada. Sensitivity (SN) and specificity (SP) were calculated for 3 scenarios: chemotherapy related ED visit, chemotherapy related H, and febrile neutropenia (FN) related visit using the chart as the gold standard. Since there is no specific diagnosis code for FN, three definitions of FN were considered: general (defined as fever or infection or neutropenia as main reason for visit), moderate (neutropenia as main reason for visit) or strict (fever or infection plus neutropenia). The population based cohort was generated by linking several health databases to identify women who had at least one ED+H during adjuvant chemotherapy for breast cancer between 2007-2009. The validation cohort consisted of 490 randomly selected cases from this cohort. Results: The population-based cohort consisted of 8,359 patients of whom 43.4% had at least one ED+H including 1,496 women who had multiple visits resulting in 6,293 unique ED+H. Of these, 73.1% were considered CRVs based on our algorithm. The algorithm performed well in identifying CRVs that included an H either from ED (SN 90%, SP 100%) or directly from home (SN 91%, SP 93%) but less well for ED visits that did not result in H (SN 65%, SP 80%). Depending on which FN algorithm was used, 1.4-24% of visits were considered FN related. The general FN algorithm had excellent SN regardless of whether the visit involved H (94-98%) but SP was moderate (66-80%). The strict FN algorithm had good SP (78-99%) but SN was highly variable (13-89%). The moderate FN algorithm provided the best tradeoff between SN (69-97%) and SP (76-98%). Conclusions: CRVs can be identified from administrative data with reasonable confidence, obviating the need for chart abstraction to evaluate chemotherapy related serious events.

Download Full-text

Can Chemotherapy-Related Acute Care Visits Be Accurately Identified in Administrative Data?

Journal of Oncology Practice ◽

10.1200/jop.2017.023697 ◽

2018 ◽

Vol 14 (1) ◽

pp. e51-e58 ◽

Cited By ~ 4

Author(s):

Monika K. Krzyzanowska ◽

Katherine Enright ◽

Rahim Moineddin ◽

Lingsong Yun ◽

Melanie Powis ◽

...

Keyword(s):

Breast Cancer ◽

Emergency Department ◽

Adjuvant Chemotherapy ◽

Acute Care ◽

Administrative Data ◽

Validation Cohort ◽

Early Stage ◽

Early Stage Breast Cancer ◽

Algorithm Validation ◽

Ed Visits

Purpose: There is increasing interest in using administrative data to examine treatment-related complications that lead to emergency department (ED) visits or hospitalizations (H). The purpose of this study was to evaluate the reliability of billing codes for identifying chemotherapy-related acute care visits (CRVs) among women with early-stage breast cancer. Materials and Methods: The cohort was identified by using deterministically linked health databases and consisted of women who were diagnosed with early-stage breast cancer who started adjuvant chemotherapy between 2007 and 2009 in Ontario, Canada. A random sample of 496 patient cases was chosen as the validation cohort. Sensitivity (SN) and specificity (SP) were calculated for three scenarios: chemotherapy-related ED visit, chemotherapy-related H, and febrile neutropenia (FN)–related visit. For FN-related visits, three definitions were considered: general, moderate, and strict. Results: The administrative cohort consisted of 8,359 patients, 43.4% of whom had at least one ED or H, including 1,496 women who had multiple visits that resulted in 6,293 unique visits. Of these, 73.1% were considered CRVs. The algorithm performed well in identifying CRVs that included H either from ED (SN, 90%; SP, 100%) or directly from home (SN, 91%; SP, 93%), but less well for ED visits that did not result in H (SN, 65%; SP, 80%). Depending on which FN algorithm was used, 4.8% to 24% of visits were considered related. The moderate FN algorithm provided the best tradeoff between SN (69% to 97%) and SP (83% to 98%). Conclusion: Administrative data can be valuable in evaluating chemotherapy-related serious events. Algorithm validation in other cohorts is needed.

Download Full-text

Acute Care Use by Breast Cancer Patients on Adjuvant Chemotherapy in Alberta: Demonstrating the Importance of Measurement to Improving Quality

Current Oncology ◽

10.3390/curroncol28060375 ◽

2021 ◽

Vol 28 (6) ◽

pp. 4420-4431

Author(s):

Che Hsuan David Wu ◽

May Lynn Quan ◽

Shiying Kong ◽

Yuan Xu ◽

Jeffrey Q. Cao ◽

...

Keyword(s):

Breast Cancer ◽

Adjuvant Chemotherapy ◽

Cancer Patients ◽

Acute Care ◽

Population Based ◽

Breast Cancer Patients ◽

Increased Risk ◽

Institutional Differences ◽

Improvement Programs ◽

Ed Visits

Breast cancer patients receiving adjuvant chemotherapy are at increased risk of acute care use. The incidence of emergency department (ED) visits and hospitalizations (H) have been characterized in other provinces but never in Alberta. We conducted a retrospective population-based cohort study using administrative data of women with stage I-III breast cancer receiving adjuvant chemotherapy. Rates of ED and H use in the 180 days following chemotherapy initiation were determined, and logistic regression was performed to identify risk factors. We found that 47% of women receiving adjuvant chemotherapy experienced ED or H, which compared favourably to other provinces. However, Alberta had the highest rate of febrile neutropenia-related ED visits, and among the highest chemotherapy-related ED visits. The incidence of acute care use increased over time, and there were significant institutional differences despite operating under a single provincial healthcare system. Our study demonstrates the need for systematic measurement and the importance of quality improvement programs to address this gap.

Download Full-text

A population-based assessment of primary care visits during adjuvant chemotherapy for breast cancer

Current Oncology ◽

10.3747/co.24.3431 ◽

2017 ◽

Vol 24 (2) ◽

pp. 90 ◽

Cited By ~ 1

Author(s):

S.J. Bastedo ◽

M.K. Krzyzanowska ◽

R. Moineddin ◽

L. Yun ◽

K.A. Enright ◽

...

Keyword(s):

Breast Cancer ◽

Primary Care ◽

Side Effects ◽

Adjuvant Chemotherapy ◽

Acute Care ◽

Early Stage ◽

Care Service ◽

Population Based ◽

Control Subjects ◽

The Impact

Background We used administrative health data to explore the impact of primary care physician (pcp) visits on acute-care service utilization by women receiving adjuvant chemotherapy for early-stage breast cancer (ebc).Methods Our population-based retrospective cohort study examined pcp visits and acute-care use [defined as an emergency room (er) visit or hospitalization] by women diagnosed with ebc between 2007 and 2009 and treated with adjuvant chemotherapy. Multivariate regression analysis was used to identify the effect of pcp visits on the likelihood of experiencing an acute-care visit.Results Patients receiving chemotherapy visited a pcp significantly more frequently than they had before their diagnosis [relative risk (rr): 1.48; 95% confidence interval (ci): 1.44 to 1.53; p < 0.001] and significantly more frequently than control subjects without cancer (rr: 1.51; 95% ci: 1.46 to 1.57; p < 0.001). More than one third of pcp visits by chemotherapy patients were related to breast cancer or chemotherapy-related side effects. In adjusted multivariate analyses, the likelihood of experiencing an er visit or hospitalization increased in the days immediately after a pcp visit (rr: 1.92; 95% ci: 1.76 to 2.10; p < 0.001).Conclusions During chemotherapy treatment, patients visited their pcp more frequently than control subjects did, and they visited for reasons related to their breast cancer or to chemotherapy-related side effects. Visits to a pcp by patients receiving chemotherapy were associated with an increased frequency of er visits or hospitalizations in the days immediately after the pcp visit. Those results suggest an opportunity to institute measures for early detection and intervention in chemotherapy side effects.

Download Full-text

Evidence for racial/ethnic disparities in emergency department visits following breast cancer surgery among women in California: a population-based study

Breast Cancer Research and Treatment ◽

10.1007/s10549-021-06119-5 ◽

2021 ◽

Author(s):

Mary Falcone ◽

Lihua Liu ◽

Albert Farias ◽

JuanJuan Zhang ◽

Irene Kang ◽

...

Keyword(s):

Breast Cancer ◽

Emergency Department ◽

Ethnic Disparities ◽

Population Based ◽

Cancer Surgery ◽

Emergency Department Visits ◽

Breast Cancer Surgery ◽

Population Based Study ◽

Racial Ethnic

Download Full-text

Characteristics and Outcomes of Women Developing Heart Failure After Early Stage Breast Cancer Chemotherapy: A Population-Based Matched Cohort Study

Circulation Heart Failure ◽

10.1161/circheartfailure.120.008110 ◽

2021 ◽

Author(s):

Husam Abdel-Qadir ◽

Felicia Tai ◽

Ruth Croxford ◽

Peter C. Austin ◽

Eitan Amir ◽

...

Keyword(s):

Breast Cancer ◽

Heart Failure ◽

Early Stage ◽

Sequential Therapy ◽

Cardiovascular Death ◽

Population Based ◽

Emergency Department Visits ◽

Early Stage Breast Cancer ◽

Administrative Databases ◽

Breast Cancer Chemotherapy

Background: The prognosis of heart failure (HF) after early stage breast cancer (EBC) treatment with anthracyclines or trastuzumab is not well-characterized. Methods: Using administrative databases, women diagnosed with HF after receiving anthracyclines or trastuzumab for EBC in Ontario during 2007 to 2017 (the EBC-HF cohort) were categorized by cardiotoxic exposure (anthracycline alone, trastuzumab alone, sequential therapy with both agents) and matched on age with ≤3 cancer-free HF controls to compare baseline characteristics. To study prognosis after HF onset, we conducted a second match on age plus important HF prognostic factors. The cumulative incidence function was used to describe risk of hospitalization or emergency department visits (hospital presentations) for HF and cardiovascular death. Results: A total of 804 women with EBC developed HF after anthracyclines (n=312), trastuzumab (n=112), or sequential therapy (n=380); they had significantly fewer comorbidities than 2411 age-matched HF controls. After the second match, the anthracycline-HF cohort had a similar 5-year incidence of HF hospital presentations (16.5% [95% CI, 12.0%–21.7%]) as controls (17.1% [95% CI, 14.4%–20.1%]); the 5-year incidence was lower than matched controls for the trastuzumab-HF (9.7% [95% CI, 4.7%–16.9%]; controls 16.4% [95% CI, 12.1%–21.3%]; P =0.03) and sequential-HF cohorts (2.7% [95% CI, 1.4%–4.8%]; controls 10.8% [95% CI, 8.9%–13.0%]; P <0.001). At 5 years, the incidence of cardiovascular death was 2.9% (95% CI, 1.2%–5.9%) in the anthracycline-HF cohort vs. 9.5% (95% CI, 6.9%–12.6%) in controls, and 1.7% (0.6%–3.7%) for women developing HF after trastuzumab vs. 4.3% (95% CI, 3.1–5.8%) for controls. Conclusions: Women developing HF after cardiotoxic EBC chemotherapy have fewer comorbidities than cancer-free women with HF; trastuzumab-treated women who develop HF have better prognosis than matched HF controls.

Download Full-text

Identifying Prenatal Opioid Exposure in Health Administrative Data for Public Health Surveillance and Epidemiologic Research

International Journal for Population Data Science ◽

10.23889/ijpds.v5i5.1453 ◽

2020 ◽

Vol 5 (5) ◽

Author(s):

Andi Camden ◽

Teresa To ◽

Joel G Ray ◽

Tara Gomes ◽

Li Bai ◽

...

Keyword(s):

Prescription Drug ◽

Administrative Data ◽

Population Based ◽

Emergency Department Visits ◽

Prescription Data ◽

Opioid Use ◽

Health Administrative Data ◽

Maternal Characteristics ◽

Data Source ◽

The Impact

IntroductionAccurate estimation of prenatal opioid exposure (POE) is needed for population-based surveillance & research but can be challenging with health administrative data due to varying definitions & methods. Prior research has relied primarily on infant records with a diagnosis of neonatal abstinence syndrome (NAS). Objectives and Approach1) Evaluate the impact of using different definitions of maternal opioid use in the estimation of POE; 2) Investigate whether maternal characteristics vary by the type of definition used. Population-based cross-sectional study of all hospital births (N= 454,746) from 2014-2017 in Ontario, Canada. Multiple linked population-based health administrative databases were used to identify opioid-related pre- & perinatal Emergency Department visits & hospitalizations & opioid prescriptions. We examined how pre-conception & in-pregnancy maternal characteristics varied by using different approaches to ascertain POE. ResultsThere were 9624 live/still births with POE. Ascertainment of POE was highest using maternal prescription drug data (79%) & infant hospital records with NAS (45%). Maternal characteristics varied by data source used for POE ascertainment. Opioid-related health care during pregnancy identified a high-risk phenotype, contrasted with those ascertained through prescription data, with respective rates of 64% vs. 54% for social assistance, 37% vs. 12% for polydrug use, 23% vs. 6% for alcohol use, 26% vs. 19% for 3+ live births, 13% vs. 5% for victim of violence, 12% vs. 6% for involvement in criminal justice system & 64% vs. 17% for mental health & addictions hospital care. Conclusion / ImplicationsPOE ascertainment differs by health administrative data source & ability to link both across maternal records and with infant. Prescription drug data identified the highest number of opioid-exposed births and, with linked healthcare records, is useful to identify illicit opioid use & additional risk factors. Clinically meaningful differences in maternal characteristics of opioid users exist by POE ascertainment method.

Download Full-text

Identification of patients at high risk for preventable emergency department visits and inpatient admissions after starting chemotherapy: Machine learning applied to comprehensive electronic health record data.

Journal of Clinical Oncology ◽

10.1200/jco.2021.39.15_suppl.1511 ◽

2021 ◽

Vol 39 (15_suppl) ◽

pp. 1511-1511

Author(s):

Dylan J. Peterson ◽

Nicolai P. Ostberg ◽

Douglas W. Blayney ◽

James D. Brooks ◽

Tina Hernandez-Boussard

Keyword(s):

Machine Learning ◽

Emergency Department ◽

Electronic Health Record ◽

Acute Care ◽

Cancer Care ◽

Model Performance ◽

Emergency Department Visits ◽

Health Record ◽

Patients At Risk ◽

Ed Visits

1511 Background: Acute care use is one of the largest drivers of cancer care costs. OP-35: Admissions and Emergency Department Visits for Patients Receiving Outpatient Chemotherapy is a CMS quality measure that will affect reimbursement based on unplanned inpatient admissions (IP) and emergency department (ED) visits. Targeted measures can reduce preventable acute care use but identifying which patients might benefit remains challenging. Prior predictive models have made use of a limited subset of the data available in the Electronic Health Record (EHR). We hypothesized dense, structured EHR data could be used to train machine learning algorithms to predict risk of preventable ED and IP visits. Methods: Patients treated at Stanford Health Care and affiliated community care sites between 2013 and 2015 who met inclusion criteria for OP-35 were selected from our EHR. Preventable ED or IP visits were identified using OP-35 criteria. Demographic, diagnosis, procedure, medication, laboratory, vital sign, and healthcare utilization data generated prior to chemotherapy treatment were obtained. A random split of 80% of the cohort was used to train a logistic regression with least absolute shrinkage and selection operator regularization (LASSO) model to predict risk for acute care events within the first 180 days of chemotherapy. The remaining 20% were used to measure model performance by the Area Under the Receiver Operator Curve (AUROC). Results: 8,439 patients were included, of whom 35% had one or more preventable event within 180 days of starting chemotherapy. Our LASSO model classified patients at risk for preventable ED or IP visits with an AUROC of 0.783 (95% CI: 0.761-0.806). Model performance was better for identifying risk for IP visits than ED visits. LASSO selected 125 of 760 possible features to use when classifying patients. These included prior acute care visits, cancer stage, race, laboratory values, and a diagnosis of depression. Key features for the model are shown in the table. Conclusions: Machine learning models trained on a large number of routinely collected clinical variables can identify patients at risk for acute care events with promising accuracy. These models have the potential to improve cancer care outcomes, patient experience, and costs by allowing for targeted preventative interventions. Future work will include prospective and external validation in other healthcare systems.[Table: see text]

Download Full-text

Specialist Palliative Care Activity at an Acute Care Tertiary Hospital and Its Representation in Administrative Data

American Journal of Hospice and Palliative Medicine® ◽

10.1177/1049909120939861 ◽

2020 ◽

pp. 104990912093986

Author(s):

Joanne M. Stubbs ◽

Hassan Assareh ◽

Helen M. Achat ◽

Sally Greenaway ◽

Poorani Muruganantham

Keyword(s):

Palliative Care ◽

Acute Care ◽

Administrative Data ◽

Tertiary Hospital ◽

Specialist Palliative Care ◽

Diagnosis Code ◽

South Wales ◽

Initial Hospital ◽

Hospital Stays ◽

Expert Input

Objective: To quantify and examine specialist palliative care (SPC) in-hospital activity and compare it to routinely collected administrative data on palliative care (PC). Methods: All patients discharged from a large acute care tertiary hospital in New South Wales, Australia, between July 1 and December 31, 2017, were identified from the hospital’s data warehouse. Administrative data were supplemented with information from the electronic medical record for hospital stays which were assigned the PC additional diagnosis code (Z51.5); had a “palliative care” care type; or included SPC consultation. Results: Of 34 653 hospital stays, 524 were coded as receiving PC—based on care type (43%) and/or diagnosis code Z51.5 (100%). Specialist palliative care provided 1717 consultations over 507 hospital stays. Patients had 2 (median; interquartile range: 1-4) consultations during an average stay of 15.3 days (SD 15.78; median 10); the first occurred 7.0 days (SD 12.13; median 3) after admission. Of patient stays with an SPC consultation, 70% were assigned the PC Z51.5 code; 60% were referred for symptom management; 68% had cancer. One hundred forty-one patients were under a palliative specialist—either from initial hospital admission (49.6%) or later in their stay. Conclusions: Palliative care specialists provide expert input into patient management, benefitting patients and other clinicians. Administrative data inadequately capture their involvement in patient care, especially consultations, and are therefore inappropriate for reporting SPC activity. Exclusion of information related to SPC activity results in an incomplete and distorted representation of PC services and fails to acknowledge the valuable contribution made by SPC.

Download Full-text