Physician Billing Claims Research Articles

Exposure to procedural ionizing radiation and cancer risk among physicians.

Physicians in certain specialities are routinely exposed to procedural ionizing radiation. Their risk of cancer is unknown, including by cancer sub-types. To assess cancer risk among exposed physicians. This population-based case-control study was completed in Ontario, Canada, where healthcare is universal, using linkage of physician billing claims to a province-wide cancer registry. Up to five cancer-free physician controls were matched to each cancer-affected physician, by sex, and both age at and year of, entry into practice. Cumulative exposure to procedural ionizing radiation was captured by physician billing claims. Conditional logistic regression generated an odds ratio (OR) of cancer per 1000 procedures performed and as a binary exposure comparing physicians above the upper 95th percentile cumulative number of procedures (≥200) to those below this cut point. Mean (standard deviation) age of the 1265 cases and 5772 non-cancer controls was 39.7 (10.7) and 37.7 (9.0) years, and 45% and 49% were female, respectively. After a median (interquartile ranges) of 13.0 (6.9-20.4) and 12.5 (6.5-20.1) years of lookback among cases and controls, the OR of cancer was 1.02 (95% confidence interval 0.99-1.05; P = NS) per 1000 additional procedures performed. Modelling the cumulative exposure to procedures nonlinearly did not change the observed association (P > 0.40 for each). Comparing physicians above versus below the upper 95th percentile cumulative number of procedures, the OR of cancer was 1.23 (95% confidence interval 0.75-2.01, P = NS). Physician exposure to procedural ionizing radiation was not associated with a higher risk of cancer. Measures that minimize radiation exposure should continue.

Model-based algorithms to ascertain smoking in administrative health data: a registry-based validation study

Objectives We developed a machine-learning model-based algorithm (MBA) for smoking in Administrative Health Data (AHD). The validity of this MBA was compared to a rule-based algorithm (RBA). Approaches The study included adults (≥18 years) from a clinical registry containing self-reported current smoking from 2017 to 2020 in Manitoba, Canada. Clinical data were linked to up to five years of hospitalization, physician billing claims, and prescription medication records. The RBA was based on diagnosis codes for tobacco use and nicotine dependence medication. MBAs, constructed using random forest (RF) models, included these indicators in addition to comorbid condition diagnoses and sociodemographic factors. Sensitivity, specificity, positive and negative predictive values (PPV, NPV), and 95% confidence intervals (CIs) were estimated. Results The cohort comprised 24,718 individuals (88.6% female); prevalence of current smokers was 10.0%. The RBA had sensitivity of 27.3% (95% CI: 24.2-30.7), specificity of 96.6% (95% CI: 96.1-97.0), and PPV of 47.2% (95% CI: 42.9-51.5). The MBA had sensitivity of 68.6% (95% CI: 65.1-71.9), specificity of 76.3% (95% CI: 75.2-77.3), and PPV of 24.3% (95% CI: 23.2-25.6). NPV was high irrespective of algorithms. Stratified analyses revealed similar estimates for males and females, and the number of years of AHD did not affect the MBA results. ConclusionsAn RF-based MBA for smoking ascertainment in linked AHD sources improved sensitivity compared to the RBA. However, the RBA excelled in specificity and PPV. ImplicationBalancing accurate smoker identification with the risk of false positives is crucial when choosing an algorithm to ascertain current smokers using AHD.

Long-Term Cardiovascular Outcomes in Children and Adolescents With Hypertension

Hypertension affects 6% of all children, and its prevalence is increasing. Childhood hypertension tracks into adulthood and is associated with subclinical cardiovascular disease; however, there is a lack of evidence linking childhood hypertension to cardiovascular outcomes, which may contribute to underdiagnosis and undertreatment. To determine the long-term associated risk of major adverse cardiac events (MACE) among children diagnosed with hypertension. This was a population-based, retrospective, matched cohort study conducted from 1996 to 2022. The study included all children (aged 3-18 years) alive in Ontario, Canada, from 1996 to 2021, who were identified using provincial administrative health databases. Children with prior kidney replacement therapy were excluded. Incident hypertension diagnosis, identified by validated case definitions using diagnostic and physician billing claims. Each case was matched with 5 controls without hypertension by age, sex, birth weight, maternal gestational hypertension, prior comorbidities (chronic kidney disease, diabetes, cardiovascular surgery), and a propensity score for hypertension. The primary outcome was MACE (a composite of cardiovascular death, stroke, hospitalization for myocardial infarction or unstable angina, or coronary intervention). Time to MACE was evaluated using the Kaplan-Meier method and Cox proportional hazards regression. A total of 25 605 children (median [IQR] age, 15 [11-17] years; 14 743 male [57.6%]) with hypertension were matched to 128 025 controls without hypertension. Baseline covariates were balanced after propensity score matching, and prior comorbidities were uncommon (hypertension vs control cohort: malignancy, 1451 [5.7%] vs 7908 [6.2%]; congenital heart disease, 1089 [4.3%] vs 5408 [4.2%]; diabetes, 482 [1.9%] vs 2410 [1.9%]). During a median (IQR) of 13.6 (7.8-19.5) years of follow-up, incidence of MACE was 4.6 per 1000 person-years in children with hypertension vs 2.2 per 1000 person-years in controls (hazard ratio, 2.1; 95% CI, 1.9-2.2). Children with hypertension were at higher associated risk of stroke, hospitalization for myocardial infarction or unstable angina, coronary intervention, and congestive heart failure, but not cardiovascular death, compared with nonhypertensive controls. Children diagnosed with hypertension had a higher associated long-term risk of MACE compared with controls without hypertension. Improved detection, follow-up, and control of pediatric hypertension may reduce the risk of adult cardiovascular disease.

Validation of case-ascertainment algorithms using health administrative data to identify people who inject drugs in Ontario, Canada

ObjectivesHealth administrative data can be used to improve the health of people who inject drugs by informing public health surveillance and program planning, monitoring, and evaluation. However, methodological gaps in the use of these data persist due to challenges in accurately identifying injection drug use (IDU) at the population level. In this study, we validated case-ascertainment algorithms for identifying people who inject drugs using health administrative data in Ontario, Canada. Study Design and SettingData from cohorts of people with recent (past 12 months) IDU, including those participating in community-based research studies or seeking drug treatment, were linked to health administrative data in Ontario from 1992 to 2020. We assessed the validity of algorithms to identify IDU over varying look-back periods (ie, all years of data [1992 onwards] or within the past 1–5 years), including inpatient and outpatient physician billing claims for drug use, emergency department (ED) visits or hospitalizations for drug use or injection-related infections, and opioid agonist treatment (OAT). ResultsAlgorithms were validated using data from 15,241 people with recent IDU (918 in community cohorts and 14,323 seeking drug treatment). An algorithm consisting of ≥1 physician visit, ED visit, or hospitalization for drug use, or OAT record could effectively identify IDU history (91.6% sensitivity and 94.2% specificity) and recent IDU (using 3-year look back: 80.4% sensitivity, 99% specificity) among community cohorts. Algorithms were generally more sensitive among people who inject drugs seeking drug treatment. ConclusionValidated algorithms using health administrative data performed well in identifying people who inject drugs. Despite their high sensitivity and specificity, the positive predictive value of these algorithms will vary depending on the underlying prevalence of IDU in the population in which they are applied.

Identifying patients with diagnosed cirrhosis in administrative health databases: a validation study.

Case ascertainment algorithms were developed and validated to identify people living with cirrhosis in administrative health data in Manitoba, Canada using primary care electronic medical records (EMR) to define the reference standards. We linked provincial administrative health data to primary care EMR data. The validation cohort included 116,675 Manitobans aged >18 years with at least one primary care visit between April 1998 and March 2015. Hospital records, physician billing claims, vital statistics, and prescription drug data were used to develop and test 93 case-finding algorithms. A validated case definition for primary care EMR data was the reference standard. We estimated sensitivity, specificity, positive and negative predictive values (PPV, NPV), Youden's index, area under the receiver operative curve, and their 95% confidence intervals (CIs). A total of 116,675 people were in the validation cohort. The prevalence of cirrhosis was 1.4% (n = 1593). Algorithm sensitivity estimates ranged from 32.5% (95% CI 32.2-32.8) to 68.3% (95% CI 68.0-68.9) and PPV from 17.4% (95% CI 17.1-17.6) to 23.4% (95% CI 23.1-23.6). Specificity (95.5-98.2) and NPV (approximately 99%) were high for all algorithms. The algorithms had slightly higher sensitivity estimates among men compared with women, and individuals aged ≥45 years compared to those aged 18-44 years. Cirrhosis algorithms applied to administrative health data had moderate validity when a validated case definition for primary care EMRs was the reference standard. This study provides algorithms for identifying diagnosed cirrhosis cases for population-based research and surveillance studies.

Estimating the completeness of physician billing claims for diabetes case ascertainment: a multiprovince investigation.

Previous research has suggested that how physicians are paid may affect the completeness of billing claims for estimating chronic disease. The purpose of this study is to estimate the completeness of physician billings for diabetes case ascertainment. We used administrative data from eight Canadian provinces covering the period 1 April 2014 to 31 March 2016. The patient cohort was stratified into two mutually exclusive groups based on their physician remuneration type: fee-for-service (FFS), for those paid only on that basis; and non-fee-for-service (NFFS). Using diabetes prescription drug data as our reference data source, we evaluated whether completeness of disease case ascertainment varied with payment type. Diabetes incidence rates were then adjusted for completeness of ascertainment. The cohort comprised 86 110 patients. Overall, equal proportions received their diabetes medications from FFS and NFFS physicians. Overall, physician payment method had little impact upon the percentage of missed diabetes cases (FFS, 14.8%; NFFS, 12.2%). However, the difference in missed cases between FFS and NFFS varied widely by province, ranging from -1.0% in Nova Scotia to 29.9% in Newfoundland and Labrador. The difference between the observed and adjusted disease incidence rates also varied by province, ranging from 22% in Prince Edward Island to 4% in Nova Scotia. The difference in the loss of cases by physician remuneration method varied across jurisdictions. This loss may contribute to an underestimation of disease incidence. The method we used could be applied to other chronic diseases for which drug therapy could serve as reference data source.

Prevalence and incidence of Paget's disease of bone: Temporal trend over 20years in the province of Quebec, Canada.

Paget's disease of bone (PDB) is a focal bone disorder characterized by an increased bone remodeling and an anarchic bone structure. A decline of prevalence and incidence of PDB has been observed in some countries. No epidemiological data are available on PDB in Canada. We aimed at examining the evolution of the prevalence and incidence of PDB in Quebec (Canada) by analyzing health administrative databases. PDB case definition relied on one or more hospitalizations, or one or more physician-billing claims with a diagnosis code of PDB. To identify incident cases, a 'run-in' period of four years (1996-1999) was used to exclude prevalent cases. For each fiscal year from 2000 to 2001 to 2019-2020 (population size 2,914,480), crude age and sex-specific prevalence and incidence rates of PDB among individuals aged ≥55years were determined, and sex-specific rates were also standardized to the 2011 age structure of the Quebec population. Generalized linear regressions were used to test for linear changes in standardized prevalence and incidence rates. Over the study period, standardized prevalence of PDB has remained stable in Quebec, from 0.44% in 2000/2001 to 0.43% in 2019/2020 (mean change -0.002, p-value=0.0935). For the 2019-2020 fiscal year, 13,165 men and women had been diagnosed with PDB and prevalence of PDB increased with age. Standardized incidence of PDB has decreased over time from 0.77/1000 in 2000/2001 to 0.28/1000 in 2019-2020 (mean change -0.228/year, p-value<0.0001), the incidence decreasing from 0.82/1000 to 0.37/1000 in men and from 0.76/1000 to 0.22/1000 in women, respectively. This decrease was observed in all age categories. With the exception of a slight increase in PDB prevalence up to 0.55% in years 2005 to 2007, the prevalence of PDB has remained stable in Quebec over the past 20years, 13,160 men and women being currently diagnosed with PDB. The incidence has decreased over time. Our results support the epidemiological changes of PDB reported in other countries.

Abstract C098: "Am I at risk for cervical cancer?": Racial disparity implications of an unexpected relationship between chronic conditions and precancerous lesions

Abstract Objective: Cervical cancer is a highly preventable malignancy caused by HPV that begins with the development of precancerous lesions (high risk cervical intraepithelial neoplasia grades 2 or 3 - hrCIN). As cancer is a metabolic disease, individuals with metabolic conditions such as diabetes, high cholesterol (hcho) and high blood pressure (hbp) may be at higher risk of developing hrCIN. Given the higher prevalence among Black women, we sought to determine whether these metabolic conditions increase the risk of hrCIN differentially by race. Methods: We used data from hospital and physician billing claims for any CIN diagnosis occurring at the Virginia Commonwealth University Health System (VCUHS) from 01/01/2014 through 12/04/2021, along with respective clinical and sociodemographic characteristics at the time of diagnosis. We estimated the risk of high-grade dysplasia (CIN2+ versus CIN1) among women with various metabolic diseases, using logistic regression (LR) with a backward selection approach (removing terms with p ≥ 0.2 and adding those with p &amp;lt; 0.1) to identify most relevant confounding factors. Results: There were 2,761 women diagnosed with CIN 1, 2, 3, and unspecified dysplasia of the cervix uteri with respective prevalence of 30%, 45%, 11%, and 14%. The prevalence of hrCIN did not vary by race (48% versus 52% for non-Latina (nL)-whites versus nL-Blacks respectively). Of the variables included in the LR model (age, marital status, religion, preferred language, insurance status, Body Mass Index, HIV, HPV, presence of bacterial vaginosis, hypertension, cholesterol, diabetes, pulmonary conditions, arthritis, heart conditions, pregnancy status, parity, smoking, alcohol, and previous cancer diagnoses), only pregnancy status, diabetes, hypertension, cholesterol, insurance, language, pulmonary conditions, and HPV remained in the final LR model. We identified an interaction between hypertension and cholesterol (p=0.03) where the odds of a hrCIN among women with both, hypertension, and cholesterol, were more than twice compared to women with neither comorbidity (OR=2.2, 95%CI: 1.1, 4.3). Stratified by race, the elevated risk of hrCIN was only significant among nL-Black women (p=0.05, OR=2.9, 95%CI: 1.1, 8.6). Conclusions: Despite no apparent difference in the incidence of hrCIN by race in this dataset, our analysis revealed an excess risk of hrCIN that was only present for Black women with both hypertension and hypercholesterolemia. These comorbidities are both overrepresented in nL-Blacks and are associated with low grade systemic inflammation, which may promote HPV persistence and progression to hrCIN. If replicable, this association deserves further investigation to assess its biological underpinnings and potential as a point of intervention to reduce the risk of CIN progression and cervical cancer. Citation Format: Shreya R. Raman, Stephanie A. Sullivan, Robert A. Winn, Katherine Y. Tossas. "Am I at risk for cervical cancer?": Racial disparity implications of an unexpected relationship between chronic conditions and precancerous lesions [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr C098.

Local Prostate Radiation Therapy and Symptomatic Local Events in De Novo Metastatic Prostate Cancer

PurposeLocal prostate radiation therapy (LPRT) for low-burden metastatic prostate cancer (mPCa) improves overall survival and is the standard of care. The role of LPRT in reducing symptomatic local events (SLE) remains unclear. We aimed to identify SLE risk factors and to evaluate the association between LPRT and SLE in mPCa. Methods and MaterialsWe conducted a retrospective, population-based cohort study of patients initially diagnosed with mPCa between 2005 and 2016 in a cancer registry. Patient, tumor, and treatment characteristics were obtained from chart review and the cancer registry. The coprimary endpoints were genitourinary (GU) and gastrointestinal (GI) SLE, identified by physician billing claims between 2004 and 2017 for diagnostic or therapeutic procedures potentially related to GU and GI SLE. The effect of LPRT on SLE was evaluated using both recurrent event (Andersen-Gill model) and time-to-first-event sequential landmark analyses. Risk factors for SLE were assessed by multivariable Cox regression. LPRT was defined as ≥40 Gy within 1 year of diagnosis. Metastatic burden was defined per the STAMPEDE trial. ResultsOf 1363 patients, 46 (3.4%) received LPRT. Median follow-up was 27.3 and 28.9 months in the control and LPRT groups, respectively. LPRT was associated with less recurrent GU SLE (hazard ratio [HR], 0.34; 95% confidence interval [CI], 0.17-0.67; P = .002), upper tract obstruction (HR, 0.20; 95% CI, 0.05-0.84; P = .03), and cystoscopy (HR, 0.38; 95% CI, 0.15-0.96; P = .04). Metastatic burden was not associated with SLE. ConclusionsLPRT in mPCa was associated with less recurrent GU SLE, specifically for upper tract obstruction and cystoscopy.

Commentary: From Mixtapes to Playlists - Evolving Options for Capturing Diagnoses in Canadian Physicians' Data.

Physician billing claims are rich sources of administrative health data. However, diagnostic codes in billing claims are drawn from the International Classification of Diseases, Ninth Revision (WHO & International Conference for the Ninth Revision of the International Classification of Diseases 1977), which has not been updated by the World Health Organization in three decades. With its updated and expanded content and its digital tooling, the International Classification of Diseases 11th Revision (ICD-11) (WHO n.d.a.) could be considered for this purpose. Primary care practitioners have always found the ICD inadequate for their needs. This may change with ICD-11, with which the International Classification of Primary Care (ICPC) (van Boven and Ten Napel 2021) is more closely aligned. ICD-11, ICPC and the Systematized Nomenclature of Medicine Clinical Terms present evolving options for capturing diagnoses in physician data.

Leaving the Walkman and ICD-9 Behind: Modernizing the Disease Classification System Used by Canadian Physicians.

The International Classification of Diseases, Ninth Revision (ICD-9) was released in the 1970s and adopted in Canada for physician billing claims in 1979 (CIHI n.d.b.; WHO & International Conference for the Ninth Revision of the International Classification of Diseases 1977). ICD-9 is no longer adequate for representing our modern healthcare environment and patient needs. We summarize the findings from a small survey of ICD-9 users across Canada – such as family physicians, researchers and decision makers – who describe the limitations of ICD-9 and the features that they would desire in a new or updated classification system.

Disparities in adherence to diabetes screening guidelines among males and females in a universal care setting: A population-based study of 1,389,697 adults.

ObjectivesNational guidelines recommend that all adults over the age of 40 years undergo screening for diabetes at least once every 3-years. We examined the association between sex and adherence to diabetes screening after accounting for age, urban/rural residence and material deprivation. We also examined the subsequent incidence of prediabetes and diabetes in adherent and non-adherent individuals. ApproachOur study is based on a retrospective population-level inception cohort of adults aged 40 - 79 years without pre-existing diabetes or cardiovascular disease on April 1, 2013. Data on hospitalizations, emergency department visits, ambulatory clinic visits, physician billing claims, pharmaceutical claims, centralized laboratory data, vital statistics death registry, and census data at the neighbourhood level were linked at the patient level for years 2013 to 2020. Adherence during a 3-year screening period (2013 – 2016) and prediabetes and diabetes during a 4-year follow-up period were examined. Multivariate logistic regression was used to examine the adjusted association between sex and adherence. ResultsAmong 1,389,697 individuals (49.2% male, 50.8% female) adherence rates were 69.9% in males and 79.8% in females. Sex-differences in adherence were largest in younger individuals (58.0% and 72.6% and in males and females aged 40-44 years, respectively) and consistent across rural/urban residence and material deprivation categories. Females were more adherent (adjusted odds ratio 1.93; 95% confidence interval 1.90 to 1.97) than males. Among adherent males and females, 65.3% and 64.8% had at least one Haemoglobin A1C, respectively during the screening period. During the follow-up period, prediabetes and diabetes rates among males who were adherent were 25.9% and 5.2%, compared to 8.8% and 2.1% among non-adherent males; and 20.2% and 2.9% among adherent and 7.3% and 1.3% among non-adherent females. ConclusionsDespite lower rates of adherence to screening, males have higher rates of prediabetes and diabetes compared to females. Our study highlights the need to develop education campaigns and targeted interventions at the local and population level to raise awareness and improve diabetes screening rates in young people, especially males.

Quality of end-of-life care for women with gynecologic malignancies in Ontario, Canada: A 14-year population-based retrospective analysis (132)

Objectives: Tracking population-based quality indicators of supportive and aggressive end of life (EOL) care may provide insight into the success of EOL interventions and health resource utilization. We examined patterns of EOL care in patients with gynecologic cancers over 14 years in Ontario, Canada. Methods: We conducted a population-based, retrospective cohort study of gynecologic cancer decedents in Ontario from 2005 to 2018 using linked administrative healthcare data (Ontario Cancer Registry, Discharge Abstract Database, Ontario Health Insurance Plan physician billing claims, homecare databases). Quality indices included emergency department (ED) use, admissions to hospital or intensive care unit (ICU), chemotherapy in the last 14 days of life, physician or palliative care home visits, and death in hospital, as well as previously defined aggregate measures of aggressive or supportive care. Descriptive statistics and crude incidence rates of each indicator were calculated. A multivariable logistic regression model was created to examine the relationship between sociodemographic and patient factors and receipt of aggressive and supportive care. Results: The cohort included 1 7,405 women; 51% died of ovarian cancer, 30% uterine cancer, 12% cervical cancer, and 6.5% vulvar or vaginal cancers (VVC). During the study period, death rates in hospitals decreased from 48% to 39%, while supportive care indices increased from 47% to 71%, and aggressive care indices remained stable (18%). Within the last 30 days of life, 53% were hospitalized, 5.6% were admitted to ICU, and 64% accessed palliative homecare, while within the last 14 days of life, 33% visited the ED, and 4.5% received chemotherapy. Patients with VVC demonstrated lower rates of both aggressive (11%) and supportive care (58%) compared to other cancer subtypes. On multivariable analyses, patients were significantly more likely to receive aggressive EOL care if they were <50 years of age, had a modified Deyo-Charlson Comorbidity Index score of 0, had uterine or ovarian cancer, or lived in a smaller community (<100,000) or a neighborhood in the lowest three income quintiles. Patients were less likely to receive supportive EOL care if they were ≥70 years of age, had a non-ovarian gynecologic malignancy, or lived in the lowest two income quintiles. Objectives: Tracking population-based quality indicators of supportive and aggressive end of life (EOL) care may provide insight into the success of EOL interventions and health resource utilization. We examined patterns of EOL care in patients with gynecologic cancers over 14 years in Ontario, Canada. Methods: We conducted a population-based, retrospective cohort study of gynecologic cancer decedents in Ontario from 2005 to 2018 using linked administrative healthcare data (Ontario Cancer Registry, Discharge Abstract Database, Ontario Health Insurance Plan physician billing claims, homecare databases). Quality indices included emergency department (ED) use, admissions to hospital or intensive care unit (ICU), chemotherapy in the last 14 days of life, physician or palliative care home visits, and death in hospital, as well as previously defined aggregate measures of aggressive or supportive care. Descriptive statistics and crude incidence rates of each indicator were calculated. A multivariable logistic regression model was created to examine the relationship between sociodemographic and patient factors and receipt of aggressive and supportive care. Results: The cohort included 1 7,405 women; 51% died of ovarian cancer, 30% uterine cancer, 12% cervical cancer, and 6.5% vulvar or vaginal cancers (VVC). During the study period, death rates in hospitals decreased from 48% to 39%, while supportive care indices increased from 47% to 71%, and aggressive care indices remained stable (18%). Within the last 30 days of life, 53% were hospitalized, 5.6% were admitted to ICU, and 64% accessed palliative homecare, while within the last 14 days of life, 33% visited the ED, and 4.5% received chemotherapy. Patients with VVC demonstrated lower rates of both aggressive (11%) and supportive care (58%) compared to other cancer subtypes. On multivariable analyses, patients were significantly more likely to receive aggressive EOL care if they were <50 years of age, had a modified Deyo-Charlson Comorbidity Index score of 0, had uterine or ovarian cancer, or lived in a smaller community (<100,000) or a neighborhood in the lowest three income quintiles. Patients were less likely to receive supportive EOL care if they were ≥70 years of age, had a non-ovarian gynecologic malignancy, or lived in the lowest two income quintiles.

Abstract 2208: Unexpected pattern in estimated prevalence of precancerous cervical lesions at a safety net hospital in Virginia: A case for epidemiologic surveillance

Abstract Objectives: High-grade squamous lesions of the cervix, (Cervical Intraepithelial Neoplasia (CIN) grade 2 and 3), are concerning, common, and costly precancerous findings, particularly for safety net hospitals who provide healthcare regardless of finances or insurance status. Epidemiologic surveillance of these lesions is uncommon but might be useful to guide preventive interventions. We sought to estimate the prevalence of CIN 2/3 and its racial and ethnic distribution, accounting for patient’s clinical and sociodemographic characteristics. Methods: We used data from hospital and physician billing claims for any CIN diagnosis occurring at the Virginia Commonwealth University Health System (VCUHS) from 01/01/2016 through 12/31/2020, along with respective clinical and sociodemographic characteristics at the time of diagnosis. We estimated average adjusted prevalence of CIN 2/3 lesions by race and ethnicity, from average marginal effects, after selecting most relevant patient factors from a logistic regression (LR) using backward selection (removing terms with p ≥ 0.2 and adding those with p &amp;lt; 0.1). Results: There were 1,538 women diagnosed with CIN 1, 2, 3, and unspecified dysplasia of the cervix uteri with respective prevalence of 45%, 14%, 12%, and 30%. CIN2/3 prevalence (26%) varied by race and ethnicity: Asians (39%) &amp;gt; Latinas (32%) &amp;gt; Whites (28%) &amp;gt;Blacks (23%) &amp;gt; Other (20%). Of the variables included in the LR model (age, marital status, religion, preferred language, insurance status, Body Mass Index, HIV, HPV, presence of bacterial vaginosis, hypertension, cholesterol, diabetes, pulmonary conditions, arthritis, heart conditions, pregnancy status, parity, smoking, alcohol, and previous cancer diagnoses), only age, diabetes, pulmonary conditions, smoking, and heart disease remained in the final LR model. The estimated average adjusted prevalence of a CIN 2/3 diagnosis remained highest for Asian women (35%, 95%CI: 17-52%), followed by Latinx (28%, 95%CI: 19-37%), Whites (25%, 95%CI 22-29%), Blacks (23%, 95%CI: 20-27%), and other (18%, 95%CI: 9-26%). Conclusions: The higher CIN 2/3 prevalence estimates among Asian women relative to other races and ethnicities is unexpected, as it does not align with current distribution of cervical cancer diagnoses at VCUHS (only 3% Asian), nor with the patterns of cervical cancer incidence and mortality rates in Virginia and in the United States, where Asian women have the lowest incidence and mortality. However, it potentially suggests a higher rate of regression, a finding supported by at least one surveillance study with similar findings. Citation Format: Shreya R. Raman, Stephanie A. Sullivan, Katherine Y. Tossas. Unexpected pattern in estimated prevalence of precancerous cervical lesions at a safety net hospital in Virginia: A case for epidemiologic surveillance [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2208.

Transitions between versions of the International Classification of Diseases and chronic disease prevalence estimates from administrative health data: a population-based study

BackgroundDiagnosis codes in administrative health data are routinely used to monitor trends in disease prevalence and incidence. The International Classification of Diseases (ICD), which is used to record these diagnoses, have been updated multiple times to reflect advances in health and medical research. Our objective was to examine the impact of transitions between ICD versions on the prevalence of chronic health conditions estimated from administrative health data.MethodsStudy data (i.e., physician billing claims, hospital records) were from the province of Manitoba, Canada, which has a universal healthcare system. ICDA-8 (with adaptations), ICD-9-CM (clinical modification), and ICD-10-CA (Canadian adaptation; hospital records only) codes are captured in the data. Annual study cohorts included all individuals 18 + years of age for 45 years from 1974 to 2018. Negative binomial regression was used to estimate annual age- and sex-adjusted prevalence and model parameters (i.e., slopes and intercepts) for 16 chronic health conditions. Statistical control charts were used to assess the impact of changes in ICD version on model parameter estimates. Hotelling’s T2 statistic was used to combine the parameter estimates and provide an out-of-control signal when its value was above a pre-specified control limit.ResultsThe annual cohort sizes ranged from 360,341 to 824,816. Hypertension and skin cancer were among the most and least diagnosed health conditions, respectively; their prevalence per 1,000 population increased from 40.5 to 223.6 and from 0.3 to 2.1, respectively, within the study period. The average annual rate of change in prevalence ranged from -1.6% (95% confidence interval [CI]: -1.8, -1.4) for acute myocardial infarction to 14.6% (95% CI: 13.9, 15.2) for hypertension. The control chart indicated out-of-control observations when transitioning from ICDA-8 to ICD-9-CM for 75% of the investigated chronic health conditions but no out-of-control observations when transitioning from ICD-9-CM to ICD-10-CA.ConclusionsThe prevalence of most of the investigated chronic health conditions changed significantly in the transition from ICDA-8 to ICD-9-CM. These results point to the importance of considering changes in ICD coding as a factor that may influence the interpretation of trend estimates for chronic health conditions derived from administrative health data.

Physician Reimbursement for Endoscopic Submucosal Dissection: A Single Center Analysis

Background and AimsEndoscopic Submucosal Dissection (ESD) has become an established procedure for resection of pre-malignant lesions. Despite its benefits and cost-effectiveness, there are significant challenges to bring ESD to mainstream practice given the unpredictability of reimbursement. ESD is submitted for billing purposes without dedicated Current procedural terminology codes. We analyzed physician component billing and reimbursement data indexed to time, procedure location, and insurance type. MethodsAll physician billing claims and reimbursement data for ESD cases performed between January 2017 and December 2019 in a single tertiary referral center were included. Procedure type was grouped by lesion location: esophageal ESD, gastric and/or duodenal ESD, and colon and/or rectal ESD. To normalize the data, we divided the total dollar payment by the number of hours it took to complete the procedure. ResultsCompleted insurance claims were available for 163 procedures, and 134 met criteria for this study. In our analysis there were 42 government, 92 private insurance payers. The mean charge was $3562.31 for government claims and $3471.33 for private claims. The mean payment was $905.01 (range: $55.64-6140) for government payers, and $1134.24 (range: $44.26-3500) for private payers. The average payment per procedure was $886.43 for esophageal ESD, $732.71 for gastric or duodenal ESD and $1325.35 for colon and/or rectal ESD. The mean $ amount for claims paid for all endoscopic mucosal resection performed in the same time period was $405.50 (range: $106-$1775). ConclusionESD is consistently under-reimbursed across insurance types when compared to surgical alternatives and endoscopist time. With proven efficacy, safety and cost-saving data, appropriate Current procedural terminology codes should be provided for ESD.

1211. Incidence of All-Cause Community-Acquired Pneumonia in Ontario and British Columbia, Canada, 2002-2018; a Canadian Immunization Research Network (CIRN) study

BackgroundCommunity-acquired pneumonia (CAP) causes substantial morbidity and mortality. There is a lack of data on the comprehensive burden of CAP across the life span in Canada. We estimated the incidence of all-cause CAP in all age groups in Ontario and British Columbia (BC), Canada. MethodsWe identified hospitalized and outpatient CAP episodes from the Discharge Database (DAD) and physician billing claims databases (Ontario Health Insurance Plan in Ontario and Medical Services Plan in BC) in both provinces. The National Ambulatory Care Reporting System was used to identify CAP episodes from emergency department visits in Ontario. CAP recorded with a primary or secondary diagnosis was identified using International Classification of Diseases 9 (480–486, 510, 513) and 10 (J10.0, J11.0, J12–J18, J86.9, J85.1) codes. We estimated the age and sex adjusted annual incidence of CAP overall, and by age groups (0–4, 5–17, 18–39, 40–64, 65–74, 75–84 and ≥85 years) according to routine childhood pneumococcal conjugate vaccine (PCV) immunization periods from 2005–2018 in Ontario and from 2002–2018 in BC. Poisson regression models were fitted with population denominators from Statistics Canada to estimate the incidence rates. ResultsOntario had 3,607,186 CAP episodes from 2005–2015 with a mean annual incidence of 2,801 (95% confidence interval [CI]: 2,748, 2,854) per 100,000 population; incidence declined from 3,077/100,000 in 2005 to 2,604/100,000 in 2010 before increasing to 2,843/100,000 in 2018. BC had 1,146,172 CAP episodes from 2002–2008, with a mean annual incidence of 2,146 (95% CI: 2105, 2189); the incidence increased from 2,005 /100,000 in 2002 to 2,199/100,000 in 2018. A high incidence of CAP was observed in children aged 0–4 years and older adults, particularly in adults aged ≥85 years in both provinces across all PCV program periods (Figure 1). Figure 1: Age group-specific incidence of all-cause community-acquired pneumonia according to childhood pneumococcal conjugate vaccine (PCV) program periods in Ontario (PCV7 [1 Jan 2005–30 Sep 2009]), PCV10 [1 Oct 2009–31 Oct 2010] and PCV13 [1 Nov 2010–31 Dec 2018]) and British Columbia (PCV7 [1 Sep 2003–31 May 2010] and PCV13 [1 Jun 2010–31 Dec 2018]), CanadaConclusionCAP continues to be a public health burden in Canada despite publicly funded pneumococcal vaccination programs. Ontario seems to have higher CAP burden than British Columbia that warrants further investigation. The youngest cohort of children and older adults contribute significantly to the CAP burden.Disclosures Manish Sadarangani, BM BCh, DPhil, GlaxoSmithKline (Grant/Research Support)Merck (Grant/Research Support)Pfizer (Grant/Research Support)Sanofi Pasteur (Grant/Research Support)Seqirus (Grant/Research Support)Symvivo (Grant/Research Support)VBI Vaccines (Research Grant or Support) Allison McGeer, MSc,MD,FRCPC,FSHEA, GlaxoSmithKline (Advisor or Review Panel member)Merck (Advisor or Review Panel member, Research Grant or Support)Pfizer (Grant/Research Support, Scientific Research Study Investigator, Advisor or Review Panel member) James D. Kellner, MD, FRCPC, FIDSA, Pfizer, Merck, GSK, Moderna (Grant/Research Support) Shaun Morris, MD, MPH, DTM&H, FRCPC, FAAP, GSK (Speaker’s Bureau)Pfizer (Advisor or Review Panel member)Pfizer (Grant/Research Support) Shaza A. Fadel, PhD MPH, Merck (Other Financial or Material Support, Salary is paid by the University of Toronto via a donation by Merck to the Centre for Vaccine Preventable Diseases to support educational and operational activities.) Fawziah Marra, BSc(Pharm), PharmD, Pfizer Canada (Research Grant or Support)

Determining the cancer diagnostic interval using administrative data in a cohort of patients with pancreatic cancer.

336 Background: Pancreatic cancer is a leading cause of cancer death, largely due to vague presenting symptoms and late stage at diagnosis. Population-based administrative data can be a valuable resource for studying the diagnostic interval. The objective of this study was to determine the first encounter in the diagnostic interval and to calculate that interval in a cohort of patients with pancreatic cancer using an empirical approach. Methods: This is a retrospective, cohort study of patients diagnosed with pancreatic ductal adenocarcinoma (PDAC) from 2007 – 2015 in Alberta, Canada. We used the Alberta Cancer Registry (ACR), physician billing claims, hospital discharge and emergency room visits to identify health encounters that occurred more frequently in the 3 months prior to diagnosis compared to those in the 3-24 months prior to diagnosis. We used statistical control charts to define the lookback period for each encounter category and identify the earliest encounter that represented the start of the diagnostic interval (index contact date). The end of the interval was the diagnosis date. Quantile regression was used to determine factors associated with the diagnostic interval. Results: We identified 3142 patients with PDAC. Median age of diagnosis was 71 (IQR 61-80). We identified an index contact date in 96.5% of the patients. The median length of the diagnostic interval was 76 days (IQR 21-191; 90th percentile 276 days). A higher Elixhauser comorbidity score (+18.57 days/ 1 point increase, 95% CI 16.07-21.07, p &lt; 0.001) and stage 3 disease (+22.55 days, 95% CI 5.02-40.08, p = 0.01) was associated with a longer diagnostic interval. Conclusions: In this cohort of patients with pancreatic cancer, there was a wide range in the diagnostic interval with 10% of patients having a diagnostic interval approaching one year. Diagnostic interval research using administrative databases can understand variations in diagnosis times, can inform early detection efforts and can improve quality of care.

Determining the cancer diagnostic interval using administrative data in a cohort of patients with pancreatic ductal adenocarcinoma (PDAC).

e13551 Background: PDAC is a leading cause of cancer death that is often diagnosed at an advanced stage. Population-based administrative data can be a valuable resource for studying the diagnostic interval, defined as the time from the first related healthcare encounter to cancer diagnosis. The objective of this study was to determine the diagnostic interval in a cohort of patients with PDAC using an empirical approach. Methods: This is a retrospective, cohort study of patients diagnosed with PDAC from 2007 – 2015 in Alberta, Canada. We used the Alberta Cancer Registry, physician billing claims, hospital discharge and emergency room visits to identify and categorize cancer-related healthcare encounters before diagnosis. We used statistical control charts to define the lookback period for each encounter category and used these lookback periods to identify the earliest encounter that represented the start of the diagnostic interval (index contact date). The end of the interval was the diagnosis date. Quantile regression was used to determine factors associated with the diagnostic interval. Results: We identified 3,142 patients with PDAC. Median age of diagnosis was 71 (IQR 61-80). We identified an index contact date and thus a diagnostic interval in 96.5% of patients. The median diagnostic interval length was 76 days (IQR 21-191; 90th percentile 276 days). A higher Elixhauser comorbidity score (+18.57 days/ 1 point increase, 95% CI 16.07-21.07, p&lt;0.001) and stage 3 disease compared to stage 2 disease (+22.55 days, 95% CI 5.02-40.08, p=0.01) were associated with a longer diagnostic interval. Conclusions: In this cohort of patients with PDAC, there was a wide range in the diagnostic interval with 10% of patients having a diagnostic interval of approximately 9 months. Diagnostic interval research using administrative databases can understand variations in diagnosis times and can inform early detection efforts by identifying where and in whom delays may occur.

A population\u2010based study to develop juvenile arthritis case definitions for administrative health data using model\u2010based dynamic classification

BackgroundPrevious research has shown that chronic disease case definitions constructed using population-based administrative health data may have low accuracy for ascertaining cases of episodic diseases such as rheumatoid arthritis, which are characterized by periods of good health followed by periods of illness. No studies have considered a dynamic approach that uses statistical (i.e., probability) models for repeated measures data to classify individuals into disease, non-disease, and indeterminate categories as an alternative to deterministic (i.e., non-probability) methods that use summary data for case ascertainment. The research objectives were to validate a model-based dynamic classification approach for ascertaining cases of juvenile arthritis (JA) from administrative data, and compare its performance with a deterministic approach for case ascertainment.MethodsThe study cohort was comprised of JA cases and non-JA controls 16 years or younger identified from a pediatric clinical registry in the Canadian province of Manitoba and born between 1980 and 2002. Registry data were linked to hospital records and physician billing claims up to 2018. Longitudinal discriminant analysis (LoDA) models and dynamic classification were applied to annual healthcare utilization measures. The deterministic case definition was based on JA diagnoses in healthcare use data anytime between birth and age 16 years; it required one hospitalization ever or two physician visits. Case definitions based on model-based dynamic classification and deterministic approaches were assessed on sensitivity, specificity, and positive and negative predictive values (PPV, NPV). Mean time to classification was also measured for the former.ResultsThe cohort included 797 individuals; 386 (48.4 %) were JA cases. A model-based dynamic classification approach using an annual measure of any JA-related healthcare contact had sensitivity = 0.70 and PPV = 0.82. Mean classification time was 9.21 years. The deterministic case definition had sensitivity = 0.91 and PPV = 0.92.ConclusionsA model-based dynamic classification approach had lower accuracy for ascertaining JA cases than a deterministic approach. However, the dynamic approach required a shorter duration of time to produce a case definition with acceptable PPV. The choice of methods to construct case definitions and their performance may depend on the characteristics of the chronic disease under investigation.