Migraineurs were reliably identified using administrative data

Association between migraine and exclusive breastfeeding: A cross-sectional study

BACKGROUND

Medical conditions may preclude a mother from exclusively breastfeeding her infant; however, the association between migraine and the duration of exclusive breastfeeding is not well known.

OBJECTIVE

To evaluate the association between migraine and the duration of exclusive breastfeeding in a representative sample of Canadian females.

METHODS

We used the Canadian Community Health Survey, a cross-sectional survey, to identify females aged 20-49 years who delivered a baby in the previous 5 years. History of migraine was self-reported. Females reported if they breastfed their baby, and among those who did, they further reported the duration of exclusive breastfeeding. We evaluated the association between migraine and the rate of breastfeeding, and the duration of exclusive breastfeeding adjusting for selected covariates.

RESULTS

We included 5282 females, of whom 862 (16.3%) had migraine. Compared to females without migraine, females with migraine were less likely to have high income (annual income >$80,000: 362 [42.0] vs. 2276 [51.6]), and more likely to have comorbid mood (176 [20.5] vs. 378 [8.6%]) and anxiety (196 [22.8%] vs. 406 [9.2%]) disorders. Migraine was not associated with breastfeeding (proportion of females who did not breastfeed, migraine vs. no migraine: 114/862 [13.2%] vs. 498/4420 [11.3%]; adjusted odds ratio 1.03; 0.74-1.27); however, females with migraine had lower odds (≥6 months of exclusive breastfeeding: 216/688 [31.4%] vs. 1325/3561 [37.2%]; adjusted odds ratio from ordinal shift analyses 0.84; 0.71-0.99) of longer duration of exclusive breastfeeding than females without migraine.

CONCLUSION

Females with migraine exclusively breastfeed their infants for a shorter duration compared to females without migraine, suggesting the need to better support this population through education on the safety and benefits of exclusive breastfeeding and better access to safe and effective treatment of migraine in lactating females.

Migraine Treatment and Healthcare Resource Utilization in Alberta, Canada

BACKGROUND

Migraine poses a significant burden worldwide; however, there is limited evidence as to the burden in Canada. This study examined the treatment patterns, healthcare resource use (HRU), and costs among newly diagnosed or recurrent patients with migraine in Alberta, Canada, from the time of diagnosis or recurrence.

METHODS

This retrospective observational study utilized administrative health data from Alberta, Canada. Patients were included in the Total Migraine Cohort if they had: (1) ≥1 International Classification of Diseases diagnostic code for migraine; or (2) ≥1 prescription dispense(s) for triptans from April 1, 2012, to March 31, 2018, with no previous diagnosis or dispensation code from April 1, 2010, to April 1, 2012.

RESULTS

The mean age of the cohort (n = 199,931) was 40.0 years and 72.3% were women. The most common comorbidity was depression (19.7%). In each medication class examined, less than one-third of the cohort was prescribed triptans and fewer than one-fifth was prescribed a preventive. Among patients with ≥1 dispense, the mean rate of opioid prescriptions was 4.61 per patient-year, compared to 2.28 triptan prescriptions per patient-year. Migraine-related HRU accounted for 3%-10% of all use.

CONCLUSION

Comorbidities and high all-cause HRU were observed among newly diagnosed or recurrent patients with migraine. There is an underutilization of acute and preventive medications in the management of migraine. The high rate of opioid use reinforces the suboptimal management of migraine in Alberta. Migraine management may improve by educating healthcare professionals to optimize treatment strategies.

Epidemiology and treatment utilization for Canadian patients with migraine: a literature review

The objective of this narrative review was to identify real-world evidence regarding the burden of migraine in Canada. We conducted a literature search in MEDLINE, Embase, and the Cochrane Database of Systematic Reviews for studies published between August 2010 and August 2020. Of the 3269 publications identified, 29 studies were included. Prevalence estimates varied widely across Canada, and mental health comorbidities were common. Individuals with migraine have a lower quality of life, detrimental impact on workforce productivity, and higher rates of health care resource utilization (HCRU), with HCRU and costs highest among those with chronic migraine. We found inconsistencies in care, including underutilization of medications such as triptans, and varied utilization of over-the-counter and prescription medications. Increased medication use was identified among those with chronic migraine, and only a small number of patients used migraine preventive medications. The burden of migraine in Canada is substantial. Reduced quality of life and workforce productivity, increased HCRU and costs, and underutilization of triptans and migraine preventive medications highlight an important need for more effective management of individuals with migraine.

Claims-based algorithms for common chronic conditions were efficiently constructed using machine learning methods

Identification of medical conditions using claims data is generally conducted with algorithms based on subject-matter knowledge. However, these claims-based algorithms (CBAs) are highly dependent on the knowledge level and not necessarily optimized for target conditions. We investigated whether machine learning methods can supplement researchers' knowledge of target conditions in building CBAs. Retrospective cohort study using a claims database combined with annual health check-up results of employees' health insurance programs for fiscal year 2016-17 in Japan (study population for hypertension, N = 631,289; diabetes, N = 152,368; dyslipidemia, N = 614,434). We constructed CBAs with logistic regression, k-nearest neighbor, support vector machine, penalized logistic regression, tree-based model, and neural network for identifying patients with three common chronic conditions: hypertension, diabetes, and dyslipidemia. We then compared their association measures using a completely hold-out test set (25% of the study population). Among the test cohorts of 157,822, 38,092, and 153,608 enrollees for hypertension, diabetes, and dyslipidemia, 25.4%, 8.4%, and 38.7% of them had a diagnosis of the corresponding condition. The areas under the receiver operating characteristic curve (AUCs) of the logistic regression with/without subject-matter knowledge about the target condition were .923/.921 for hypertension, .957/.938 for diabetes, and .739/.747 for dyslipidemia. The logistic lasso, logistic elastic-net, and tree-based methods yielded AUCs comparable to those of the logistic regression with subject-matter knowledge: .923-.931 for hypertension; .958-.966 for diabetes; .747-.773 for dyslipidemia. We found that machine learning methods can attain AUCs comparable to the conventional knowledge-based method in building CBAs.

The association between asthma and perinatal mental illness: a population-based cohort study

BACKGROUND

Asthma is a risk factor for mental illness, but few studies have explored this association around the time of pregnancy. We studied the association between asthma and perinatal mental illness and explored the modifying effects of social and medical complexities.

METHODS

In a population-based cohort of 846 155 women in Ontario, Canada, with a singleton live birth in 2005-2015 and no recent history of mental illness, modified Poisson regression models were constructed to examine the association between asthma diagnosed before pregnancy and perinatal mental illness, controlling for socio-demographics and medical history. We explored the modifying effects of social and medical complexities using relative excess risk due to interaction. Additional analyses examined the association between asthma and perinatal mental illness by timing and type of mental illness.

RESULTS

Women with asthma were more likely than those without asthma to have perinatal mental illness [adjusted relative risk (aRR) 1.14; 95% (confidence interval) CI: 1.13, 1.16]. Asthma was associated with increased risk of diagnosis of mental illness prenatally (aRR 1.11; 95% CI: 1.08, 1.13) and post-partum (aRR 1.17; 95% CI: 1.15, 1.19) and specifically diagnoses of mood and anxiety disorders (aRR 1.14; 95% CI: 1.13, 1.16), psychotic disorders (aRR 1.20; 95% CI: 1.10, 1.31) and substance- or alcohol-use disorders (aRR 1.24; 95% CI: 1.14, 1.36). There was no effect modification related to social or medical complexity for these outcomes.

CONCLUSIONS

Women with asthma predating pregnancy are at slightly increased risk of mental illness in pregnancy and post-partum. A multidisciplinary management strategy may be required to ensure timely identification and treatment.

Case-Ascertainment Models to Identify Adults with Obstructive Sleep Apnea Using Health Administrative Data: Internal and External Validation

Background

There is limited evidence on whether obstructive sleep apnea (OSA) can be accurately identified using health administrative data.

Study Design and Methods

We derived and validated a case-ascertainment model to identify OSA using linked provincial health administrative and clinical data from all consecutive adults who underwent a diagnostic sleep study (index date) at two large academic centers (Ontario, Canada) from 2007 to 2017. The presence of moderate/severe OSA (an apnea–hypopnea index≥15) was defined using clinical data. Of 39 candidate health administrative variables considered, 32 were tested. We used classification and regression tree (CART) methods to identify the most parsimonious models via cost-complexity pruning. Identified variables were also used to create parsimonious logistic regression models. All individuals with an estimated probability of 0.5 or greater using the predictive models were classified as having OSA.

Results

The case-ascertainment models were derived and validated internally through bootstrapping on 5099 individuals from one center (33% moderate/severe OSA) and validated externally on 13,486 adults from the other (45% moderate/severe OSA). On the external cohort, parsimonious models demonstrated c-statistics of 0.75–0.81, sensitivities of 59–60%, specificities of 87–88%, positive predictive values of 79%, negative predictive values of 73%, positive likelihood ratios (+LRs) of 4.5–5.0 and –LRs of 0.5. Logistic models performed better than CART models (mean integrated calibration indices of 0.02–0.03 and 0.06–0.12, respectively). The best model included: sex, age, and hypertension at the index date, as well as an outpatient specialty physician visit for OSA, a repeated sleep study, and a positive airway pressure treatment claim within 1 year since the index date.

Interpretation

Among adults who underwent a sleep study, case-ascertainment models for identifying moderate/severe OSA using health administrative data had relatively low sensitivity but high specificity and good discriminative ability. These findings could help study trends and outcomes of OSA individuals using routinely collected health care data.

Perinatal Complications as a Mediator of the Association Between Chronic Disease and Postpartum Mental Illness

Background: Chronic disease is associated with increased risk of postpartum mental illness, but the mechanisms underlying this association are unclear. Our aim was to explore the mediating role of perinatal complications in the association between chronic disease and postpartum mental illness. Materials and Methods: This was a population-based retrospective cohort study of all women in Ontario, Canada, from 2005 to 2015 with a singleton live birth and no recent history of mental illness during or 2 years before pregnancy. The outcome was mental illness diagnosis between delivery and 365 days postpartum, with perinatal complications, including pregnancy, delivery, and neonatal complications. Modified Poisson regression models were used to examine the association between chronic disease and perinatal mental illness, with generalized estimating equations for the calculation of total, direct, and indirect effects. All models were adjusted for sociodemographic characteristics and remote history of mental health care. Results: Of the 792,972 women, 21.1% had a chronic disease. Chronic disease was associated with an increased risk of postpartum mental illness (adjusted relative risk [aRR] 1.15 [95% confidence interval, CI 1.14-1.16]). There was no evidence of an indirect effect of chronic disease on postpartum mental illness via perinatal complications (aRR 1.003, 95% CI 1.002-1.003). Perinatal complications explained only 1.5% of the association between chronic disease and postpartum mental illness. Results were consistent by type of perinatal complication and chronic disease diagnosis. Conclusion: We observed no clinically meaningful mediating effect of perinatal complications in the association between chronic disease and postpartum mental illness. Future research should investigate alternative mechanisms explaining this association.

Association Between Chronic Medical Conditions and Acute Perinatal Psychiatric Health-Care Encounters Among Migrants: A Population-Based Cohort Study

OBJECTIVES

To examine the relationship between prepregnancy chronic medical conditions (CMCs) and the risk of acute perinatal psychiatric health-care encounters (i.e., psychiatric emergency department visits, hospitalizations) among refugees, nonrefugee immigrants, and long-term residents in Ontario.

METHODS

We conducted a population-based study of 15- to 49-year-old refugees (N = 29,189), nonrefugee immigrants (N = 187,430), and long-term residents (N = 641,385) with and without CMC in Ontario, Canada, with a singleton live birth in 2005 to 2015 and no treatment for mental illness in the 2 years before pregnancy. Modified Poisson regression was used to estimate the relative risk of a psychiatric emergency department visit or hospitalization from conception until 1 year postpartum among women with versus without CMC, stratified by migrant status. An unstratified model with an interaction term between CMC and migrant status was used to test for multiplicativity of effects.

RESULTS

The association between CMC and risk of a psychiatric emergency department visit or hospitalization was stronger among refugees (adjusted relative risk [aRR] = 1.87; 95% confidence interval [CI], 1.36 to 2.58) compared to long-term residents (aRR = 1.39; 95% CI, 1.30 to 1.48; interaction P = 0.047). The strength of the association was no different in nonrefugee immigrants (aRR = 1.26; 95% CI, 1.05 to 1.51) compared to long-term residents (interaction P = 0.45).

CONCLUSION

Our study identifies refugee women with CMC as a high-risk group for acute psychiatric health care in the perinatal period. Preventive psychosocial interventions may be warranted to enhance supportive resources for all women with CMC and, in particular refugee women, to reduce the risk of acute psychiatric health care in the perinatal period.

Bootstrap imputation minimized misclassification bias when measuring Colles' fracture prevalence and its associations using health administrative data

OBJECTIVES

Misclassification bias can result from the incorrect assignment of disease status using inaccurate diagnostic codes in health administrative data. This study quantified misclassification bias in the study of Colles' fracture.

STUDY DESIGN AND SETTING

Colles' fracture status was determined in all patients >50 years old seen in the emergency room at a single teaching hospital between 2006 and 2014 by manually reviewing all forearm radiographs. This data set was linked to population-based data capturing all emergency room visits. Reference disease prevalence and its association with covariates were measured. A multivariate model using covariates derived from administrative data was used to impute Colles' fracture status and measure its prevalence and associations using bootstrapping methods. These values were compared with reference values to measure misclassification bias. This was repeated using diagnostic codes to determine Colles' fracture status.

RESULTS

Five hundred eighteen thousand, seven hundred forty-four emergency visits were included with 3,538 (0.7%) having a Colles' fracture. Determining disease status using the diagnostic code (sensitivity 69.4%, positive predictive value 79.9%) resulted in significant underestimate of Colles' fracture prevalence (relative difference -13.3%) and biased associations with covariates. The Colles' fracture model accurately determined disease probability (c-statistic 98.9 [95% confidence interval {CI} 98.7-99.1], calibration slope 1.009 [95% CI 1.004-1.013], Nagelkerke's R² 0.71 [95% CI 0.70-0.72]). Using disease probability estimates from this model, bootstrap imputation (BI) resulted in minimal misclassification bias (relative difference in disease prevalence -0.01%). The statistical significance of the association between Colles' fracture and age was accurate in 32.4% and 70.4% of samples when using the code or BI, respectively.

CONCLUSION

Misclassification bias in estimating disease prevalence and its associations can be minimized with BI using accurate disease probability estimates.

Bootstrap imputation with a disease probability model minimized bias from misclassification due to administrative database codes

OBJECTIVE

Diagnostic codes used in administrative databases cause bias due to misclassification of patient disease status. It is unclear which methods minimize this bias.

STUDY DESIGN AND SETTING

Serum creatinine measures were used to determine severe renal failure status in 50,074 hospitalized patients. The true prevalence of severe renal failure and its association with covariates were measured. These were compared to results for which renal failure status was determined using surrogate measures including the following: (1) diagnostic codes; (2) categorization of probability estimates of renal failure determined from a previously validated model; or (3) bootstrap methods imputation of disease status using model-derived probability estimates.

RESULTS

Bias in estimates of severe renal failure prevalence and its association with covariates were minimal when bootstrap methods were used to impute renal failure status from model-based probability estimates. In contrast, biases were extensive when renal failure status was determined using codes or methods in which model-based condition probability was categorized.

CONCLUSION

Bias due to misclassification from inaccurate diagnostic codes can be minimized using bootstrap methods to impute condition status using multivariable model-derived probability estimates.