Breaking the 80:20 rule in health research using large administrative data sets

Objective: To explore the application of online analytic processing (OLAP) to improve the efficiency of analytics using large administrative health data sets. Methods: 18 years of administrative health data (1994/95 to 2012/13) were obtained from the Alberta Ministry of Health in Canada. The data sets included hospitalization, ambulatory care and practitioner claims data. Reference files were obtained that provided information including patient demographics, resident postal code, facility, and provider details. Population counts and projections for each year, sex, age were included for rate calculations. These sources were used to develop a data cube using OLAP tools. Results: Time required for analyses was reduced to 5% of that required when comparing run-time for simple queries that did not require linkage of data sets. The data cube negated the need for many intermediary steps for data extraction and analyses for research activities. Conventional methods required over 250 GB of server space for multiple analytic subsets, compared to only 10.3 GB for the data cube. Conclusions: Cross-training in information technology and health analytics is recommended to provide capacity to better leverage OLAP tools which are available with many common applications.

Opioid prescribing among new users for non-cancer pain in the USA, Canada, UK, and Taiwan: A population-based cohort study

BACKGROUND

The opioid epidemic in North America has been driven by an increase in the use and potency of prescription opioids, with ensuing excessive opioid-related deaths. Internationally, there are lower rates of opioid-related mortality, possibly because of differences in prescribing and health system policies. Our aim was to compare opioid prescribing rates in patients without cancer, across 5 centers in 4 countries. In addition, we evaluated differences in the type, strength, and starting dose of medication and whether these characteristics changed over time.

METHODS AND FINDINGS

We conducted a retrospective multicenter cohort study of adults who are new users of opioids without prior cancer. Electronic health records and administrative health records from Boston (United States), Quebec and Alberta (Canada), United Kingdom, and Taiwan were used to identify patients between 2006 and 2015. Standard dosages in morphine milligram equivalents (MMEs) were calculated according to The Centers for Disease Control and Prevention. Age- and sex-standardized opioid prescribing rates were calculated for each jurisdiction. Of the 2,542,890 patients included, 44,690 were from Boston (US), 1,420,136 Alberta, 26,871 Quebec (Canada), 1,012,939 UK, and 38,254 Taiwan. The highest standardized opioid prescribing rates in 2014 were observed in Alberta at 66/1,000 persons compared to 52, 51, and 18/1,000 in the UK, US, and Quebec, respectively. The median MME/day (IQR) at initiation was highest in Boston at 38 (20 to 45); followed by Quebec, 27 (18 to 43); Alberta, 23 (9 to 38); UK, 12 (7 to 20); and Taiwan, 8 (4 to 11). Oxycodone was the first prescribed opioid in 65% of patients in the US cohort compared to 14% in Quebec, 4% in Alberta, 0.1% in the UK, and none in Taiwan. One of the limitations was that data were not available from all centers for the entirety of the 10-year period.

CONCLUSIONS

In this study, we observed substantial differences in opioid prescribing practices for non-cancer pain between jurisdictions. The preference to start patients on higher MME/day and more potent opioids in North America may be a contributing cause to the opioid epidemic.

Informing Primary Care Changes in Alberta: Continuity and Potential Impacts on Acute Care

Health systems across Canada are embarking on initiatives to enhance access to primary care services, with the intent of improving patient outcomes and mitigating escalating healthcare costs. However, it is important that such initiatives be carefully weighed with the evidence that the changes will indeed have the desired impact. In Alberta, part of the informative process involved an analysis to examine links between continuity with primary care and utilization of acute care services. The findings provide information regarding expectations for outcomes and potentially useful (and not so useful) measures for monitoring progress and performance.

A big data analytics platform to support simulation modeling for osteoarthritis care pathways

IntroductionTechnical solutions have been used in industry settings for many years to facilitate efficient management and analyses of big data sources. An initiative to apply a business solution to support development of simulation models for health systems research using nearly two decades of provincial administrative health data is described. Objectives and ApproachAdministrative data including practitioner claims, hospitalizations and ambulatory care visits for patients with a diagnosis of osteoarthritis were obtained from Alberta Health for the period 1994/95 to 2012/13. These data were incorporated into a multidimensional data cube using Microsoft SQL Server Analysis Services. Initial steps required dimensional modeling to restructure the data into a star schema format. This involved appending several data sets and defining additional reference tables to contain stratification variables and denominator data for rate calculations. The modeling expert worked closely with the information technology team throughout the process and assessed validity of the output. ResultsDevelopment and validation of the multidimensional cube occurred in iterations over approximately 12 months. The final solution resulted in an analytics platform that compiled data from approximately 400 million records obtained from four different administrative data sources. Ten dimension tables containing 102 variables provided enhanced flexibility to conduct ad hoc stratified analyses in a fraction of the time that would be required using conventional methods. For example, some analyses that previously required a day of analyst time could be performed in less than 15 minutes. The efficiencies in analytic time were achieved by the pre-aggregated measures and slice and dice capability of the data cube, which negated many intermediary steps for data extraction and time consuming iterative analyses required for development of the simulation models. Conclusion/ImplicationsThis project demonstrated how a technical solution applied in industry can be utilized to address challenges encountered by researchers related to managing and analyzing large administrative health data sets. The methods could be applied in many other research settings to facilitate access to and analyses of information using big data.

Applications of Big Data Analytics within a Dynamic Simulation Modeling Platform to Inform Osteoarthritis Care in Alberta

IntroductionOsteoarthritis (OA) is a leading cause of chronic disability. There is need to leverage administrative data to support OA policy analysis. Our objective was to develop and apply a multidimensional data cube as an input parameter repository using health administrative data to populate an OA simulation model. Objectives and ApproachHealth administrative data including practitioner claims, inpatient and ambulatory visits from 1994 to 2013 were integrated into a multidimensional data cube. OA cases were identified using validated algorithms, and followed through stages of care (primary, specialist, acute and post-operative). The cube provided rate calculations, duration and average cost for each stage of care across the model dimensions (age categories, sex, comorbidity status and geographic zones). The rates were then linked to the model as input parameters to simulate patient flow across the continuum of care. We used the model to predict direct costs across all dimensions from 2010 to 2035. ResultsUsing the model, total number of patients with OA in Alberta will increase from 312,000 in 2010 to 1.4 million in 2035. The average annual cost per OA patient also increases from $2,800 to $4,900, and the total cost increases from $450 million in 2010 to 2.2 billion in 2035. The majority of the patients were at earlier stages (non-surgical 78%, surgical 22%), with lower average cost (non-surgical $3,300 vs. surgical $16,400) in 2010. As new administrative data are being provided routinely, the data cube is capable of providing real-time updates for the input parameters of the model, which will aid in validation of the model results and improving the precision of projections. Conclusion/ImplicationsThe data cube has significantly improved our ability to manage and analyze administrative data within a simulation model to project the burden of OA in Alberta. The integrated model can be used as a real time decision-support tool to inform osteoarthritis service planning and variations in resource utilization.

Impact of diet on mortality from stroke: results from the U.S. multiethnic cohort study

OBJECTIVES

Stroke is the fourth leading cause of death in the United States and stroke mortality rates vary by ethnicity. The purpose of this study was to examine the associations between food group consumption and risk of death from stroke among 5 ethnic groups in the United States.

METHODS

The Multiethnic Cohort includes >215,000 participants, the majority of whom are African American, Native Hawaiian, Japanese American, Latino, and Caucasian men and women recruited by mail survey in Hawaii and Los Angeles in 1993-1996. Deaths from stroke were identified by linkage to the state death files and the U.S. National Death Index. Diet was assessed using a validated food frequency questionnaire. Associations were examined using multivariable Cox proportional hazards models, stratified by ethnicity and gender.

RESULTS

A total of 860 deaths from stroke were identified among the cohort participants. Vegetable intake was associated with a significant reduction in risk for fatal stroke among African American women (relative risk [RR]=0.60; 95% CI: 0.36-0.99). Among Japanese American women only, high fruit intake was significantly associated with a risk reduction for stroke mortality (RR=0.43; 95% confidence interval [CI]: 0.22-0.85), whereas meat intake increased risk (RR=2.36; 95% CI: 1.31-4.26). Among men, a significant reduction in stroke mortality was observed among Native Hawaiians (RR=0.26; 95% CI: 0.07-0.95). After pooling the data for the ethnic groups, the findings support an elevated risk for high meat intake among women overall (RR=1.56; 95% CI: 1.12-2.16); no significant effects of dietary intake on risk for fatal stroke were observed among men.

CONCLUSIONS

Although some variations were observed for the associations between diet and stroke mortality among ethnic groups, the findings suggest that these differences are not substantial and may be due to dietary intake of specific food subgroups. Additional investigations including dietary subgroups and nutrients sources are needed to clarify these findings.

Fruit and vegetable consumption, ethnicity and risk of fatal ischemic heart disease

OBJECTIVE

Mortality rates from ischemic heart disease vary among ethnic groups. Dietary intake of fruits and vegetables has been associated with a lower risk of ischemic heart disease, but ethnic-specific data are limited.

DESIGN

Prospective cohort study.

SETTING

Hawaii and Los Angeles County, between 1993 and 1996.

PARTICIPANTS

These analyses included 164,617 adults age 45 to 75, representing five ethnic groups who were enrolled in the Multiethnic Cohort Study. Dietary data were collected at baseline using a validated food frequency questionnaire and fatal ischemic heart disease cases were identified up to December 31, 2001. Associations between fruit and vegetable consumption and fatal ischemic heart disease were examined using multivariate Cox proportional hazard models.

RESULTS

The associations between fruit and vegetable intake and fatal ischemic heart disease were similar among the five ethnic groups. When data for the ethnic groups were combined, higher vegetable intake was associated with a protective effect against ischemic heart disease in men with all intake levels above 3.4 servings per day (over 6.6 servings per day: hazard ratio, 0.73; 95% confidence interval, 0.58-0.92). Inconsistent results were observed for women, where the protective association was observed only at mid-level vegetable intake levels, but not among women with the highest level of vegetable intake. There was no evidence of an association for fruit intake.

CONCLUSIONS

Associations between fruit and vegetable intake and fatal IHD do not appear to vary among ethnic groups. Additional research is needed to clarify associations for fruit versus vegetable intake and impact on cardiovascular outcomes.

The nature of the copper atoms of cytochrome c oxidase as studied by optical and x-ray absorption edge spectroscopy

X-ray absorption edge spectroscopy has been used to study the copper of 1--2 mM cytochrome c oxidase in the resting oxidized, mixed-valence, and fully reduced states. A comparison was made of this protein with copper complexes and with natural and artificial copper proteins. Spectra were obtained with synchrotron radiation from the SPEAR storage ring using highly sensitive fluorescence detectors. Temperatures of -80 to -120 degrees C were employed further to improve the stability of the samples and to avoid the possibility of either auto- or photon-induced reduction of the materials, which might have occurred in previous studies. In order to characterize the valence states of the Cu and Fe components, the samples were monitored by infrared and visible spectroscopy before and after irradiation by the X-ray beam. The combination of the optical and X-ray absorption techniques has afforded a deconvolution of the four species of copper in the various states of cytochrome c oxidase and the tentative assignment of Cu alpha, the copper redox coupled to the heme alpha of cytochrome alpha, as a highly covalent type of copper and Cu alpha 3, the copper of cytochrome alpha 3, as a more ionic 'blue' type I copper. The implications of these findings upon the mechanism of action of cytochrome oxidase are briefly outlined.