Clinical Events and Specialty Agreement: Multi-Site, Anomaly Detection, Longitudinal Analysis
| dc.contributor | Patient-Centered Outcomes Research Institute |
| dc.contributor.author | PEDSnet Data Coordinating Center |
| dc.contributor.other | PEDSnet Data Coordinating Center |
| dc.date.accessioned | 2024-09-09T17:01:56Z |
| dc.date.created | 2024-06-05 |
| dc.description.abstract | This check highlights anomalous types of clinical events that occur across clinician or facility specialties and across time. It provides a high level summary of the quality of clinical specialty data in a study sample to examine multiple sites. The intention of the check is to evaluate the concordance of specific events (e.g., diagnosis of type I diabetes) and specialist providers or care sites (e.g., endocrinologist or endocrinology clinic) at the visit level. |
| dc.identifier.uri | https://hdl.handle.net/20.500.14642/761 |
| dc.identifier.uri | https://doi.org/10.24373/pdsp-456 |
| dc.publisher | PEDSnet |
| dc.rights | a CC-BY Attribution 4.0 License. |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0 |
| dc.subject | Multi-Site Analysis |
| dc.subject | Data Anomaly Method |
| dc.subject | Longitudinal Analysis |
| dc.subject | Visit-Level Analysis |
| dc.title | Clinical Events and Specialty Agreement: Multi-Site, Anomaly Detection, Longitudinal Analysis |
| dspace.entity.type | DQCheck |
| local.code.github | https://github.com/ssdqa/clinicalevents.specialties |
| local.code.package | # install.packages("devtools") devtools::install_github('ssdqa/https://github.com/ssdqa/clinicalevents.specialties') |
| local.description.raw | The raw data output requires the user to annotate the specialty name output before running the output function. <br> | Column | Data Type | Definition | |------------------------|---------------------|---------------------------------------------------------------------------------------------------------| |`specialty_name` | character | *user provided annotation* | |`specialty_concept_id` | numeric / character | the concept indicating the specialty of the provider and/or care site | |`specialty_concept_name` | character | if a vocabulary table is provided, the name of the specialty associated with the `specialty_concept_id` | {.dqcheck-table} <br> The resulting raw data output of this check produces nine columns of data: <br> | Column | Data Type | Definition | |----------------------|---------------------|------------------------------------------------------------------------------------------------------------------| |`specialty_concept_id` | numeric / character | the concept indicating the specialty of the provider and/or care site | |`cluster` | character | the cluster assigned in the user provided codeset; if no cluster is indicated, the name of the concept is used | |`num_visits` | numeric | the number of visits associated with the `specialty_concept_id`and cluster | |`codeset_name` | character | the name of the codeset provided for the analysis | |`site` | character | the name of the site being targeted OR "combined" if multiple sites were provided for a single site analysis | |`time_start` | date | the start of the time period being examined | |`time_increment` | character | the length of each time period | |`age_grp` | character | if an age group dataframe is provided, a string indicating the age grouping at the time of the specialty visit | |`visit_type` | character | if a visit type dataframe is provided, the visit type of the specialty visit | {.dqcheck-table} |
| local.description.viz | This check outputs three visualizations to display the Euclidean distance between two time series: the smoothed (Loess) proportion of a user-selected variable for a given site, and the average proportion of all sites. Two line graphs (one smoother, one raw) represent the proportion of the variable at each site over time. Sites are differentiated by color, and a thick red line represente the All Site Average. A circular bar graph displays the Euclidean distance from the all-site mean where the color represents the average Loess proportion over time. |
| local.dqcheck.category | Consistency |
| local.dqcheck.clinicalprobe | Targeted Patient Population |
| local.dqcheck.clinicalprobe | Utilization Patterns |
| local.dqcheck.clinicalprobe | Clinical Consistency |
| local.dqcheck.clinicalprobe | Confirmatory Clinical Data |
| local.dqcheck.measurement | Euclidean Distance |
| local.dqcheck.probe | Missing Expected Data |
| local.dqcheck.probe | External Benchmarking |
| local.dqcheck.probe | Temporality Consistency Check |
| local.dqcheck.probe | Anomalous Values from Internal Distributions |
| local.dqcheck.requirement | cohort |
| local.dqcheck.requirement | multi_or_single_site |
| local.dqcheck.requirement | age_groups |
| local.dqcheck.requirement | codeset_tbl |
| local.dqcheck.requirement | care_site |
| local.dqcheck.requirement | provider |
| local.dqcheck.requirement | visit_type_tbl |
| local.dqcheck.requirement | time |
| local.dqcheck.requirement | time_span |
| local.dqcheck.requirement | time_period |
| local.dqcheck.requirement | vocab_tbl |
| local.dqcheck.type | Dataset Fitness |
| local.dqcheck.viz | Line Graph |
| local.dqcheck.viz | Radial Bar Graph |
| relation.isCodeOfDQCheck | 929c8dfc-2c8b-4e62-8e1d-0fa06c542832 |
| relation.isCodeOfDQCheck.latestForDiscovery | 929c8dfc-2c8b-4e62-8e1d-0fa06c542832 |
| relation.isDQResultOfDQCheck | c08bdf1d-ad36-4620-ba7b-46ddc0809d48 |
| relation.isDQResultOfDQCheck.latestForDiscovery | c08bdf1d-ad36-4620-ba7b-46ddc0809d48 |