Best Practices for Binary and Ordinal Data Analyses.

The measurement of many human traits, states, and disorders begins with a set of items on a questionnaire. The response format for these questions is often simply binary (e.g., yes/no) or ordered (e.g., high, medium or low). During data analysis, these items are frequently summed or used to estimate factor scores. In clinical applications, such assessments are often non-normally distributed in the general population because many respondents are unaffected, and therefore asymptomatic. As a result, in many cases these measures violate the statistical assumptions required for subsequent analyses. To reduce the influence of the non-normality and quasi-continuous assessment, variables are frequently recoded into binary (affected-unaffected) or ordinal (mild-moderate-severe) diagnoses. Ordinal data therefore present challenges at multiple levels of analysis. Categorizing continuous variables into ordered categories typically results in a loss of statistical power, which represents an incentive to the data analyst to assume that the data are normally distributed, even when they are not. Despite prior zeitgeists suggesting that, e.g., variables with more than 10 ordered categories may be regarded as continuous and analyzed as if they were, we show via simulation studies that this is not generally the case. In particular, using Pearson product-moment correlations instead of maximum likelihood estimates of polychoric correlations biases the estimated correlations towards zero. This bias is especially severe when a plurality of the observations fall into a single observed category, such as a score of zero. By contrast, estimating the ordinal correlation by maximum likelihood yields no estimation bias, although standard errors are (appropriately) larger. We also illustrate how odds ratios depend critically on the proportion or prevalence of affected individuals in the population, and therefore are sub-optimal for studies where comparisons of association metrics are needed. Finally, we extend these analyses to the classical twin model and demonstrate that treating binary data as continuous will underestimate genetic and common environmental variance components, and overestimate unique environment (residual) variance. These biases increase as prevalence declines. While modeling ordinal data appropriately may be more computationally intensive and time consuming, failing to do so will likely yield biased correlations and biased parameter estimates from modeling them.

Bedside estimates of dead space using end-tidal CO2 are independently associated with mortality in ARDS.

PURPOSE: In acute respiratory distress syndrome (ARDS), dead space fraction has been independently associated with mortality. We hypothesized that early measurement of the difference between arterial and end-tidal CO2 (arterial-ET difference), a surrogate for dead space fraction, would predict mortality in mechanically ventilated patients with ARDS. METHODS: We performed two separate exploratory analyses. We first used publicly available databases from the ALTA, EDEN, and OMEGA ARDS Network trials (N = 124) as a derivation cohort to test our hypothesis. We then performed a separate retrospective analysis of patients with ARDS using University of Chicago patients (N = 302) as a validation cohort. RESULTS: The ARDS Network derivation cohort demonstrated arterial-ET difference, vasopressor requirement, age, and APACHE III to be associated with mortality by univariable analysis. By multivariable analysis, only the arterial-ET difference remained significant (P = 0.047). In a separate analysis, the modified Enghoff equation ((PaCO2-PETCO2)/PaCO2) was used in place of the arterial-ET difference and did not alter the results. The University of Chicago cohort found arterial-ET difference, age, ventilator mode, vasopressor requirement, and APACHE II to be associated with mortality in a univariate analysis. By multivariable analysis, the arterial-ET difference continued to be predictive of mortality (P = 0.031). In the validation cohort, substitution of the arterial-ET difference for the modified Enghoff equation showed similar results. CONCLUSION: Arterial to end-tidal CO2 (ETCO2) difference is an independent predictor of mortality in patients with ARDS.

Multilevel Factor Score Regression.

Multilevel SEM is an increasingly popular technique to analyze data that are both hierarchical and contain latent variables. The parameters are usually jointly estimated using a maximum likelihood estimator (MLE). This has the disadvantage that a large sample size is needed and misspecifications in one part of the model may influence the whole model. We propose an alternative stepwise estimation method, which is an extension of the Croon method for factor score regression. In this article, we extend this method to the multilevel setting. A simulation study was used to compare this new estimation method to the standard MLE. The Croon method outperformed MLE with regard to convergence rate, bias, MSE, and coverage, in particular when models contained a structural misspecification. In conclusion, the Croon method seems to be a promising alternative to MLE.

A Comparison of Pediatric Weight Estimation Methods for Emergency Resuscitation.

OBJECTIVES: Obtaining accurate pediatric weight is necessary during emergency resuscitation. Although several weight estimation methods exist, the most precise method has not been conclusively determined. This study aimed to evaluate the validity, reliability, and practicality of these tools. METHODS: A prospective observational study was conducted in healthy Thai children aged 6 months to 12 years. Correlations between estimated and actual weights were tested. Validity was assessed by mean bias (estimated weight minus actual weight) and accuracy (10% error). Practicality was evaluated by time usage and data derived from user questionnaires. RESULTS: Four hundred thirty participants with mean age of 6.7 years and mean weight of 26 kg were enrolled. A strong correlation between estimated weight and actual weight in all methods was demonstrated. Parental estimation was the most accurate tool in all age groups, with the lowest overall mean error (ME) of -0.83 kg and the highest accuracy of 88.7%. The Broselow tape was the second most accurate tool in ages younger than 1 year and 1-to-5-year age groups (ME = 0.23 and 0.50 kg; accuracy = 55.3% and 54.1%, respectively). The Mercy method was the second most accurate tool in the 6-to-10-year and 11-to-12-year age groups (ME = -2.47 and -2.77; accuracy = 54.6% and 67.9%, respectively). The Broselow tape had the highest score for practicality of use. CONCLUSIONS: Parental estimation was the most accurate method in every age group. The next best alternative is the Broselow tape in children aged 5 years or younger and the Mercy method in children aged older than 5 years.

Effect of Obesity and Underweight Status on Perioperative Outcomes of Congenital Heart Operations in Children, Adolescents, and Young Adults: An Analysis of Data From the Society of Thoracic Surgeons Database.

BACKGROUND: Extreme body mass index (BMI; either very high or very low) has been associated with increased risk of adverse perioperative outcome in adults undergoing cardiac surgery. The effect of BMI on perioperative outcomes in congenital heart disease patients has not been evaluated. METHODS: A multicenter retrospective cohort study was performed studying patients 10 to 35 years of age undergoing a congenital heart disease operation in the Society of Thoracic Surgeons Congenital Heart Surgery Database between January 1, 2010, and December 31, 2015. The primary outcomes were operative mortality and a composite outcome (1 or more of operative mortality, major adverse event, prolonged hospital length of stay, and wound infection/dehiscence). The associations between age- and sex-adjusted BMI percentiles and these outcomes were assessed, with adjustment for patient-level risk factors, with multivariate logistic regression. RESULTS: Of 18 337 patients (118 centers), 16% were obese, 15% were overweight, 53% were normal weight, 7% were underweight, and 9% were severely underweight. Observed risks of operative mortality (P=0.04) and composite outcome (P<0.0001) were higher in severely underweight and obese subjects. Severely underweight BMI was associated with increased unplanned cardiac operation and reoperation for bleeding. Obesity was associated with increased risk of wound infection. In multivariable analysis, the association between BMI and operative mortality was no longer significant. Obese (odds ratio, 1.28; P=0.008), severely underweight (odds ratio, 1.29; P<0.0001), and underweight (odds ratio, 1.39; P=0.002) subjects were associated with increased risk of composite outcome. CONCLUSIONS: Obesity and underweight BMI were associated with increased risk of composite adverse outcome independently of other risk factors. Further research is necessary to determine whether BMI represents a modifiable risk factor for perioperative outcome.

Associations between comorbidities, their treatment and survival in patients with interstitial lung diseases - a claims data analysis.

BACKGROUND: Interstitial lung diseases (ILDs) are associated with a high burden of disease. However, data on the prognostic impact of comorbidities and comorbidity-related pharmaceutical treatments in patients with various ILDs remain sparse. METHODS: Using longitudinal claims data from a German Statutory Health Insurance Fund, we assessed comorbidity in ILD subtypes and associated drug treatments. Baseline comorbidity was assessed via the Elixhauser Comorbidity Index that was amended by ILD-relevant conditions. Drug treatment was assessed on the substance level using the ATC-codes of drugs prescribed at the time of ILD diagnosis. Subsequently, the comorbid conditions (main analysis) and pharmaceutical substances (secondary analysis) with a meaningful association to survival were identified for the complete ILD cohort and within the subtype strata. For this, we applied multivariate Cox models using a LASSO selection process and visualized the findings within comorbidomes. RESULTS: In the 36,821 patients with ILDs, chronic obstructive pulmonary disease (COPD), arterial hypertension, and ischaemic heart disease (IHD) were the most prevalent comorbidities. The majority of patients with cardiovascular diseases received pharmaceutical treatment, while, in other relevant comorbidities, treatment quotas were low (COPD 46%, gastro-oesophageal reflux disease 65%). Comorbidities had a clinically meaningful detrimental effect on survival that tended to be more pronounced in the case of untreated conditions (e.g. hazard ratios for treated IHD 0.97 vs. 1.33 for untreated IHD). Moreover, comorbidity impact varied substantially between distinct subtypes. CONCLUSIONS: Our analyses suggest that comorbid conditions and their treatment profile significantly affect mortality in various ILDs. Therefore, comprehensive comorbidity assessment and management remains important in any ILD.

Prevalence of asthma, aspirin sensitivity and allergy in chronic rhinosinusitis: data from the UK National Chronic Rhinosinusitis Epidemiology Study.

BACKGROUND: Chronic rhinosinusitis (CRS) is a common disorder associated with other respiratory tract diseases such as asthma and inhalant allergy. However, the prevalence of these co-morbidities varies considerably in the existing medical literature and by phenotype of CRS studied. The study objective was to identify the prevalence of asthma, inhalant allergy and aspirin sensitivity in CRS patients referred to secondary care and establish any differences between CRS phenotypes. METHODS: All participants were diagnosed in secondary care according to international guidelines and invited to complete a questionnaire including details of co-morbidities and allergies. Data were analysed for differences between controls and CRS participants and between phenotypes using chi-squared tests. RESULTS: The final analysis included 1470 study participants: 221 controls, 553 CRS without nasal polyps (CRSsNPs), 651 CRS with nasal polyps (CRSwNPs) and 45 allergic fungal rhinosinusitis (AFRS). The prevalence of asthma was 9.95, 21.16, 46.9 and 73.3% respectively. The prevalence of self-reported confirmed inhalant allergy was 13.1, 20.3, 31.0 and 33.3% respectively; house dust mite allergy was significantly higher in CRSwNPs (16%) compared to CRSsNPs (9%, p < 0.001). The prevalence of self- reported aspirin sensitivity was 2.26, 3.25, 9.61 and 40% respectively. The odds ratio for aspirin sensitivity amongst those with AFRS was 28.8 (CIs 9.9, 83.8) p < 0.001. CONCLUSIONS: The prevalence of asthma and allergy in CRS varies by phenoytype, with CRSwNPs and AFRS having a stronger association with both. Aspirin sensitivity has a highly significant association with AFRS. All of these comorbidities are significantly more prevalent than in non-CRS controls and strengthen the need for a more individualised approach to the combined airway.

A Generic Model for Follicular Lymphoma: Predicting Cost, Life Expectancy, and Quality-Adjusted-Life-Year Using UK Population-Based Observational Data.

OBJECTIVES: To use real-world data to develop a flexible generic decision model to predict cost, life expectancy, and quality-adjusted life-years (QALYs) for follicular lymphoma (FL) in the general patient population. METHODS: All patients newly diagnosed with FL in the UK's population-based Haematological Malignancy Research Network (www.hmrn.org) between 2004 and 2011 were followed until 2015 (N = 740). Treatment pathways, QALYs, and costs were incorporated into a discrete event simulation to reflect patient heterogeneity, including age and disease management. Two scenario analyses, based on the latest National Institute for Health and Clinical Excellence (NICE) guidelines (rituximab induction therapy for newly diagnosed asymptomatic patients and rituximab maintenance therapy for patients between treatments), were conducted and their economic impacts were compared to current practice. RESULTS: Incidence-based analysis revealed expected average lifetime costs ranging from £6,165 [US$7,709] to £63,864 [US$79,862] per patient, and average life expectancy from 75 days to 17.56 years. Prevalence-based analysis estimated average annual treatment costs of £60-65 million [US$75-80 million], accounting for approximately 10% of the United Kingdom's annual National Health Service budget for hematological cancers as a whole. Assuming that treatment effects reported in trials are applicable to all patient groups, scenario analyses for two recent NICE guidelines demonstrated potential annual cost savings for the United Kingdom that ranged with uptake frequency from £0.6 million to £11 million [US$0.75-2.75 million]. CONCLUSIONS: Costs, survival, and QALYs associated with FL vary markedly with patient characteristics and disease management. Allowing the production of more realistic outcomes across the patient population as a whole, our model addresses this heterogeneity and is a useful tool with which to evaluate new technologies/treatments to support healthcare decision makers.

Incidence of in-hospital and post-discharge diagnosed hospital-associated venous thromboembolism using linked administrative data.

BACKGROUND: Hospital-associated venous thromboembolism (HA-VTE) is a serious adverse event, preventable with appropriate care during and post-admission. Accurate measurement of in-hospital and post-discharge incidences is essential for implementation and evaluation of prevention strategies and monitoring. AIMS: To estimate in-hospital and post-discharge diagnosed VTE, trends and risk factors. METHODS: This was a population-based study in New South Wales, Australia, using linked hospital admission and emergency department data for 2010-2013 of adult patients with a minimum stay of 48 h. HA-VTE were diagnosed in-hospital or post-discharge (within 90 days). Multi-level modelling schemes produced adjusted rates and ratios for patient, admission and hospital-related characteristics. RESULTS: From 1 865 059 admissions, the HA-VTE incidence rate was 9.7 per 1000 admissions; 71% were diagnosed post-discharge, and 4.3% died with a greater risk for VTE diagnosed in hospital compared to post-discharge (8.4% vs 2.6%, P < 0.001). Compared with surgical patients, medical patients developed fewer HA-VTE (IRR = 0.60, 95% CI: 0.58-0.63) but were more likely to be diagnosed post-discharge (OR = 2.19; 95% CI: 2.00-2.40). HA-VTE increased 6.5% over the period, driven by the 44% increase in in-hospital diagnoses and not by the 9% decrease in post-discharge diagnoses. CONCLUSIONS: HA-VTE is a continuing burden, and diagnosis after recent hospital discharge is notably high. Incidence varies across patients and facilities, highlighting the need for individual VTE risk assessment. Inclusive measures and routine monitoring of HA-VTE incidence and mortality are essential for implementing best practice and assessing effectiveness of prevention strategies.

Physical activity and fitness are associated with verbal memory, quality of life and depression among nursing home residents: preliminary data of a randomized controlled trial.

BACKGROUND: Few studies have simultaneously examined changes in physical, cognitive and emotional performance throughout the aging process. METHODS: Baseline data from an ongoing experimental randomized study were analyzed. Physical activity, handgrip, the Senior Fitness Test, Trail Making Test A, Rey Auditory-Verbal Learning Test, Quality of Life-Alzheimer's Disease Scale (QoL-AD) and the Goldberg Depression Scale were used to assess study participants. Logistic regression models were applied. TRIAL REGISTRATION: ACTRN12616001044415 (04/08/2016). RESULTS: The study enrolled 114 participants with a mean age of 84.9 (standard deviation 6.9) years from ten different nursing homes. After adjusting for age, gender and education level, upper limb muscle strength was found to be associated with Rey Auditory-Verbal Learning Test [EXP(B): 1.16, 95% confidence interval (CI): 1.04-1.30] and QoL-AD [EXP(B): 1.18, 95% CI: 1.06-1.31]. Similarly, the number of steps taken per day was negatively associated with the risk of depression according to the Goldberg Depression Scale [EXP(B): 1.14, 95% CI: 1.000-1.003]. Additional analyses suggest that the factors associated with these variables are different according to the need for using an assistive device for walking. In those participants who used it, upper limb muscle strength remained associated with Rey Auditory-Verbal Learning Test [EXP(B): 1.21, 95% CI: 1.01-1.44] and QoL-AD tests [EXP(B): 1.19, 95% CI: 1.02-1.40]. In those individuals who did not need an assistive device for walking, lower limb muscle strength was associated with Rey Auditory-Verbal Learning Test [EXP(B): 1.35, 95% CI: 1.07-1.69], time spent in light physical activity was associated with QoL-AD test [EXP(B): 1.13, 95% CI: 1.00-1.02], and the number of steps walked per day was negatively associated with the risk of depression according to the Goldberg Depression Scale [EXP(B): 1.27, 95% CI: 1.000-1.004]. CONCLUSIONS: Muscle strength and physical activity are factors positively associated with a better performance on the Rey Auditory-Verbal Learning Test, QoL-AD and Goldberg Depression Scale in older adults with mild to moderate cognitive impairment living in nursing homes. These associations appeared to differ according to the use of an assistive device for walking. Our findings support the need for the implementation of interventions directed to increase the strength and physical activity of individuals living in nursing homes to promote physical, cognitive and emotional benefits. TRIAL REGISTRATION: ACTRN12616001044415 (04/08/2016).

Automatic Measurement of Fetal Brain Development from Magnetic Resonance Imaging: New Reference Data.

BACKGROUND: Accurate fetal brain volume estimation is of paramount importance in evaluating fetal development. The aim of this study was to develop an automatic method for fetal brain segmentation from magnetic resonance imaging (MRI) data, and to create for the first time a normal volumetric growth chart based on a large cohort. SUBJECTS AND METHODS: A semi-automatic segmentation method based on Seeded Region Growing algorithm was developed and applied to MRI data of 199 typically developed fetuses between 18 and 37 weeks' gestation. The accuracy of the algorithm was tested against a sub-cohort of ground truth manual segmentations. A quadratic regression analysis was used to create normal growth charts. The sensitivity of the method to identify developmental disorders was demonstrated on 9 fetuses with intrauterine growth restriction (IUGR). RESULTS: The developed method showed high correlation with manual segmentation (r2 = 0.9183, p < 0.001) as well as mean volume and volume overlap differences of 4.77 and 18.13%, respectively. New reference data on 199 normal fetuses were created, and all 9 IUGR fetuses were at or below the third percentile of the normal growth chart. DISCUSSION: The proposed method is fast, accurate, reproducible, user independent, applicable with retrospective data, and is suggested for use in routine clinical practice.

Increased suicides during new year, but not during Christmas in Sweden: analysis of cause of death data 2006-2015.

BACKGROUND: Previous studies have investigated the impact of the Christmas and New Year holiday on suicide rates. However, no such data has yet been published on Swedish suicides. AIMS: To examine the occurrence of suicides on these dates in Sweden between 2006 and 2015. METHODS: The suicide count for each date between December 15th and January 15th was obtained from the Swedish cause of death registry. The observations were transformed to Z-scores to enable calculation of p-values. RESULTS: A small but non-significant decrease in suicides was observed on Christmas and New Year's Eve. A significant spike was found on New Year's Day (NYD) (Z = 3.40; p < .001), and these excess suicide occurred mainly among men aged 15-24 and 45-64. However, the number of suicides were somewhat lower than expected on the 31st of December (Z = -1.58; p = .115). DISCUSSION: The noted increase in suicide on NYD is in line with previous research from other countries. However, the decrease in suicides on the day before NYD suggests a delay rather than a spontaneous increase of suicides. Possible mechanisms to explain this phenomenon are discussed, such as the "broken promise effect", increased alcohol consumption, or lower help-seeking and accessibility to care.

Age-Period-Cohort Analysis of Stroke Mortality in China: Data From the Global Burden of Disease Study 2013.

BACKGROUND AND PURPOSE: Stroke has been the leading cause of death in China. The aim of this study is to assess the long-term trends of stroke mortality in China between 1994 and 2013. METHODS: The mortality data were obtained from the GBD 2013 (Global Burden of Disease Study 2013) and were analyzed with the age-period-cohort framework. RESULTS: We found that the net drift was -2.665% (95% confidence interval, -2.854% to -2.474%) per year for men and -4.064% (95% confidence interval, -4.279% to -3.849%) per year for women, and the local drift values were below 0 in all age groups (P<0.05 for all) in both sexes during the period of 1994 to 2013. In the same birth cohort, the risk of death from stroke rose exponentially with age for both sexes after controlling for period deviations. The estimated period and cohort relative risks were found in similar monotonic downward patterns (significantly with P<0.05 for all) for both sexes, with more quickly decreasing for women than for men during the whole period (significantly with P<0.05 for both). CONCLUSIONS: The decreased mortality rates of stroke in China are likely to be related to improvements in medical care and techniques, spectacular economic growth and fast urbanization, and better early life nutrition conditions of Chinese people. Besides, better education and better awareness of stroke-related knowledge in successive generations could also probably play a role.

Pattern and Outcome of Heart Failure-Related Hospitalization Over 5 Years in a Remote Australian Population: A Retrospective Administrative Data Cohort of 617 Indigenous and non-Indigenous Cases.

OBJECTIVE: The aim of this work was to understand the pattern and outcomes for heart failure (HF)-related hospitalization among Indigenous and non-Indigenous patients living in Central Australia. METHODS AND RESULTS: A retrospective analysis of administrative data for patients presenting with a primary or secondary diagnosis of HF to Central Australia's Alice Springs Hospital during 2008-2012 was performed. The population rate of admission and subsequent outcomes (including mortality and readmission) during the 5-year study period were examined. A total of 617 patients, aged 55.8 ± 17.5 years and 302 (49%) female constituted the study cohort. The 446 Indigenous patients (72%) were significantly younger (50.8 ± 15.9 vs 68.7 ± 14.9; P < .001) and clinically more complex compared with the non-Indigenous patients. Annual prevalence of any HF hospitalization was markedly higher in the Indigenous population (1.9%, 95% CI 1.7-2.1) compared with the non-Indigenous population (0.5%, 95% CI 0.4-0.6); the greatest difference being for women. Overall, non-Indigenous patients had poorer outcomes and were significantly more likely to die (P < .0001), but this was largely driven by age differences. Alternatively, Indigenous patients were significantly more likely to have a higher number of hospitalizations, although indigeneity was not a predictor for 30- or 365-day rehospitalization from the index admission. CONCLUSION: The pattern of HF among Indigenous Australians in Central Australia is characterized by a younger population with more clinically complex cases and greater health care utilization.

The long-term outcome of reduced-intensity allogeneic stem cell transplantation from a matched related or unrelated donor, or haploidentical family donor in patients with leukemia: a retrospective analysis of data from the China RIC Cooperative Group.

This study compared 6-year follow-up data from patients undergoing reduced-intensity conditioning (RIC) transplantation with an HLA-matched related donor (MRD), an HLA-matched unrelated donor (MUD), or an HLA-haploidentical donor (HID) for leukemia. Four hundred and twenty-seven patients from the China RIC Cooperative Group were enrolled, including 301 in the MRD, 79 in the HID, and 47 in the MUD groups. The conditioning regimen involved fludarabine combined with anti-lymphocyte globulin and cyclophosphamide. Graft-versus-host disease (GVHD) prophylaxis was administered using cyclosporin A (CsA) and mycophenolate mofetil (MMF). Four hundred and nineteen patients achieved stable donor chimerism. The incidence of stage II-IV acute GVHD in the HID group was 44.3 %, significantly higher than that in the MRD (23.6 %) and MUD (19.1 %) groups. The 1-year transplantation-related mortality (TRM) rates were 44.3, 17.6, and 21.3, respectively. Event-free survival (EFS) at 6 years in the HID group was 36.7 %, significantly lower than that of the MRD and MUD groups (59.1 and 66.0 %, P < 0.001 and P = 0.001, respectively). For advanced leukemia, the relapse rate of the HID group was 18.5 %, lower than that of the MRD group (37.5 %, P = 0.05), but the EFS at 6 years was 31.7 and 30.4 % (P > 0.05), respectively. RIC transplantation with MRD and MUD had similar outcome in leukemia which is better than that with HID. RIC transplantation with HID had lower relapsed with higher TRM and GVHD rate, particularly in advanced leukemias. RIC transplantation with MRD and MUD had similar outcomes in leukemia and they were better than those with HID. RIC transplantation with HID had a lower relapse rate but higher TRM and GVHD rates, particularly in cases of advanced leukemia.

Using prescription monitoring program data to characterize out-of-pocket payments for opioid prescriptions in a state Medicaid program.

BACKGROUND: Out-of-pocket payment for prescription opioids is believed to be an indicator of abuse or diversion, but few studies describe its epidemiology. Prescription drug monitoring programs (PDMPs) collect controlled substance prescription fill data regardless of payment source and thus can be used to study this phenomenon. OBJECTIVE: To estimate the frequency and characteristics of prescription fills for opioids that are likely paid out-of-pocket by individuals in the Oregon Medicaid program. RESEARCH DESIGN: Cross-sectional analysis using Oregon Medicaid administrative claims and PDMP data (2012 to 2013). SUBJECTS: Continuously enrolled nondually eligible Medicaid beneficiaries who could be linked to the PDMP with two opioid fills covered by Oregon Medicaid. MEASURES: Patient characteristics and fill characteristics for opioid fills that lacked a Medicaid pharmacy claim. Fill characteristics included opioid name, type, and association with indicators of high-risk opioid use. RESULTS: A total of 33 592 Medicaid beneficiaries filled a total of 555 103 opioid prescriptions. Of these opioid fills, 74 953 (13.5%) could not be matched to a Medicaid claim. Hydromorphone (30%), fentanyl (18%), and methadone (15%) were the most likely to lack a matching claim. The 3 largest predictors for missing claims were opioid fills that overlapped with other opioids (adjusted odds ratio [aOR] 1.37; 95% confidence interval [CI], 1.34-1.4), long-acting opioids (aOR 1.52; 95% CI, 1.47-1.57), and fills at multiple pharmacies (aOR 1.45; 95% CI, 1.39-1.52). CONCLUSIONS: Prescription opioid fills that were likely paid out-of-pocket were common and associated with several known indicators of high-risk opioid use.

Using Search Engine Query Data to Explore the Epidemiology of Common Gastrointestinal Symptoms.

BACKGROUND: Internet searches are an increasingly used tool in medical research. To date, no studies have examined Google search data in relation to common gastrointestinal symptoms. AIMS: The aim of this study was to compare trends in Internet search volume with clinical datasets for common gastrointestinal symptoms. METHODS: Using Google Trends, we recorded relative changes in volume of searches related to dysphagia, vomiting, and diarrhea in the USA between January 2008 and January 2011. We queried the National Inpatient Sample (NIS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS) during this time period and identified cases related to these symptoms. We assessed the correlation between Google Trends and these two clinical datasets, as well as examined seasonal variation trends. RESULTS: Changes to Google search volume for all three symptoms correlated significantly with changes to NIS output (dysphagia: r = 0.5, P = 0.002; diarrhea: r = 0.79, P < 0.001; vomiting: r = 0.76, P < 0.001). Both Google and NIS data showed that the prevalence of all three symptoms rose during the time period studied. On the other hand, the NHAMCS data trends during this time period did not correlate well with either the NIS or the Google data for any of the three symptoms studied. Both the NIS and Google data showed modest seasonal variation. CONCLUSIONS: Changes to the population burden of chronic GI symptoms may be tracked by monitoring changes to Google search engine query volume over time. These data demonstrate that the prevalence of common GI symptoms is rising over time.

A Shared Opportunity for Improving Electronic Medical Record Data.

With the recent rapid adoption of electronic medical records (EMRs), studies reporting results based on EMR data have become increasingly common. While analyzing data extracted from our EMR for a retrospective study, we identified various types of erroneous data entries. This report investigates the root causes of the incompleteness, inconsistency, and inaccuracy of the medical records analyzed in our study. While experienced health information management professionals are well aware of the many shortcomings with EMR data, the aims of this case study are to highlight the significance of the negative impact of erroneous EMR data, to provide fundamental principles for managing EMRs, and to provide recommendations to help facilitate the successful use of electronic health data, whether to inform clinical decisions or for clinical research.

Utility of Web search query data in testing theoretical assumptions about mephedrone.

OBJECTIVE: With growing access to the Internet, people who use drugs and traffickers started to obtain information about novel psychoactive substances (NPS) via online platforms. This paper aims to analyze whether a decreasing Web interest in formerly banned substances-cocaine, heroin, and MDMA-and the legislative status of mephedrone predict Web interest about this NPS. METHODS: Google Trends was used to measure changes of Web interest on cocaine, heroin, MDMA, and mephedrone. Google search results for mephedrone within the same time frame were analyzed and categorized. RESULTS: Web interest about classic drugs found to be more persistent. Regarding geographical distribution, location of Web searches for heroin and cocaine was less centralized. Illicit status of mephedrone was a negative predictor of its Web search query rates. The connection between mephedrone-related Web search rates and legislative status of this substance was significantly mediated by ecstasy-related Web search queries, the number of documentaries, and forum/blog entries about mephedrone. CONCLUSIONS: The results might provide support for the hypothesis that mephedrone's popularity was highly correlated with its legal status as well as it functioned as a potential substitute for MDMA. Google Trends was found to be a useful tool for testing theoretical assumptions about NPS.