data service center
The Center for Outcomes Research (COR) often provides data analytics and statistics services for other departments at Houston Methodist.
How to Request Data for Research
Requesting data through COR can be a two-step process. Below we have outlined the process for requesting data for research.
- The Center for Outcomes Research is under the Houston Methodist Research Institute (HMRI) Institutional Review Board (IRB) oversight for all data access and analysis involving protected health information. Thus, row-level data (identified or not) cannot be released without IRB approval. If you are just looking for an aggregate count of “how many patients met X” criteria, to decide if it is feasible to proceed with a study, we can help with that as a first step.
- Obtaining an aggregate patient count for determination of feasibility does not require an IRB protocol covering the data extraction, but we do require some detailed information from you or your principal investigator (PI) to do this properly. Complete and precisely coded inclusion and exclusion criteria (ICD-10) must be supplied by the study team. Please reach out to Dr. Stephen L. Jones, MD, MSHI for an example Excel template for providing inclusion and exclusion criterion from our team if you need one. We do not perform chart reviews. These tasks fall back to your study team.
- What we need to help you assess feasibility of a study protocol:
- Explicit inclusion criteria and explicit exclusion criteria: Explicit and precise diagnoses codes (e.g. “advanced heart failure patients” isn’t precise enough. We cannot guess what you mean, and every principal investigator (PI) has their own twist on what “advanced heart failure patients” means). Provide all diagnosis and/or procedure codes of interest for both the inclusion and exclusion criteria. If you don’t know what the various codes are, reach out to someone that handles the billing in the area where these patients receive care.
- Start Date and End Date of the period of interest. Epic went online at Houston Methodist on 05/22/2016. We do not have efficient access to data from prior to 05/22/2016 for the time being.
- Patient factors (e.g. age, gender, ethnicity etc.)
- Any required break-outs of the counts.
2. Full Research Data Sets: IRB Approval Required. If you started with a feasibility summary, most of these steps have already occurred. At this phase, we usually just need the full list of all approved variables to extract them for the analysis.
- Step 1: Prepare a complete research protocol with analytic plan, including a detailed list of variables and definitions requested for your study. Please indicate the start and stop dates for the study.
- Step 2: Submit the protocol, analytic plan and the detailed lists of variables and definitions to the Houston Methodist Institutional Review Board (IRB) for review and approval.
- Step 3: Once the protocol and data elements for the study have been approved by the IRB , contact the director of health informatics at COR, Dr. Stephen L. Jones, MD, MSHI with your data request. Please include the complete IRB approved protocol, the approval letter and the data request templates, as attachments in the email.
- After receiving the documentation requested above, the request is reviewed, enqueued, assigned a priority and assigned to a team of analysts and data engineers within the COR to pull the data for your approved study. We typically create a Microsoft Teams group to keep track of correspondence, documentation and to provide a HIPAA compliant file-share system that is fully audited, but other arrangements are also possible.
- The priority and complexity of the request dictate the turn-around time. Our priorities are set by the leadership of the HMRI and our bandwidth. Current (2021/2022) prioritization rubric:
- Funded COVID research projects
- COVID research projects of clinical or operational interest to the organization
- Extramurally funded other research projects
- Intramurally funded research projects
- All other projects
Statistical Support through COR
The Center for Outcomes Research offers statistical support by providing consultations for study design and analyses to grant/contract-supported research teams and clinical research groups through a cost recovery model. Statistical personnel specialize in research project statistical design and conceptualization, development of specific aims and hypotheses, statistical power and sample size determination, data analyses plans, missing data analytic procedures, inferential hypothesis testing with parametric/non-parametric tests, univariate and multivariate regression modeling with diagnostics and multicollinearity assessment, generalized linear models, categorical data analysis, repeated measurements analysis, time-to-event survival data analyses, longitudinal data analysis, multilevel/mixed model analysis, hierarchical modeling, multivariate analysis, factor analysis, nonparametric modeling and analysis, data cleaning, coding and merging, psychometrics, survey design, survey data analysis and data mining. For statistical software, Stata is used for the bulk of analytic analyses, and PASS is used for power and sample size determination. COR has also acquired a license to the geographic information system ArcGIS desktop software (by Esri) for mapping and analytics.
In general, we use three models when providing statistical services to other research teams in other departments:
- Tier 1: Collaborative Projects - Internal pilot projects, externally funded projects, and grant applications
- Tier 2: Faculty Development - Statistical support for junior clinical faculty
- Tier 3: Fee-for-service for non-integrative research requests
Center for Outcomes Research Policy on Authorship for Statistical Services
Given that the COR staff member has contributed to the research in a substantial way, charging for statistical services does not exclude COR faculty and staff from authorship on manuscripts. Ideally, determination for authorship will be established at the beginning of the project so that both researchers are aware of each other’s criteria. As project goals and expectations change, the authorship agreement can be revisited.
The Center for Outcomes Research uses the current International Committee of Medical Journal Editors (ICMJE) recommendations regarding authorship.
The ICMJE recommends that authorship should be “based on the following 4 criteria:
- Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND
- Drafting the work or revising it critically for important intellectual content; AND
- Final approval of the version to be published; AND
- Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.”
Given that these 4 criteria are met, COR faculty/staff should be invited to be a co-author on the manuscript, regardless if the researcher is charged for services or not.