Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada

Abstract Background Hospitalizations are a resource-intensive form of healthcare use, particularly for persons with chronic conditions such as HIV. In standardized Canadian hospitalization databases, it can be unclear whether a hospitalization record is an independent hospitalization, a planned inte...

Full description

Saved in:
Bibliographic Details
Main Authors: SD Emerson, T McLinden, P Sereda, J Trigg, AM Yonkman, KA Salters, S Au, KW Kooij, MO Budu, VD Lima, R Barrios, RS Hogg
Format: Article
Language:English
Published: BMC 2025-07-01
Series:BMC Medical Research Methodology
Online Access:https://doi.org/10.1186/s12874-025-02602-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849334699789385728
author SD Emerson
T McLinden
P Sereda
J Trigg
AM Yonkman
KA Salters
S Au
KW Kooij
MO Budu
VD Lima
R Barrios
RS Hogg
author_facet SD Emerson
T McLinden
P Sereda
J Trigg
AM Yonkman
KA Salters
S Au
KW Kooij
MO Budu
VD Lima
R Barrios
RS Hogg
author_sort SD Emerson
collection DOAJ
description Abstract Background Hospitalizations are a resource-intensive form of healthcare use, particularly for persons with chronic conditions such as HIV. In standardized Canadian hospitalization databases, it can be unclear whether a hospitalization record is an independent hospitalization, a planned interhospital transfer, or an unplanned readmission. Misclassifying hospitalization records can bias metrics (e.g., counting transfers as readmissions can inflate readmission counts) and hence yield incorrect results. We compared definitions for combining sequential, related hospitalization records to create hospitalization episodes of care (HEoC) within a cohort of persons with and without HIV (PWH; PWoH) in British Columbia (BC), Canada. Methods Acute care hospitalization records (April 1992 to March 2020) were sourced from the Discharge Abstract Database within the Comparative Outcomes And Service Utilization Trends (COAST) study, a BC data linkage that includes samples of PWH and PWoH. Guided by published approaches and data quality considerations, we compared eight HEoC definitions applied to PWH and PWoH. Definitions varied by the date gap between records (0 day [same-day] or ≤ 1 day), and transfer indication (none required, populated transfer fields, one-way matching of hospital transfer identifiers, or two-way matching of hospital transfer identifiers). Comparisons were primarily informed by the percentage of multi-record HEoCs (HEoCs involving multiple hospitalization records, including interhospital transfers), and feasibility given data quality. Results The sample included 56,455 hospitalization records from 10,826 PWH, and 973,430 hospitalization records from 299,053 PWoH. Across the eight HEoC definitions, the percentage of multi-record HEoCs varied from 2.8 to 6.0% among PWH and 3.6 to 5.5% among PWoH. Definitions yielding the highest percentage of multi-record HEoCs combined records without requiring a transfer indication; definitions yielding the lowest percentage of multi-record HEoCs required two-way agreement of hospital identifiers. Patterns were generally comparable among PWH and PWoH, and similar in sensitivity analyses. Conclusions Various approaches can be used to define HEoCs. We recommended a balanced HEoC definition – requiring at least one populated hospital identifier field (without requiring matching of hospital identifiers) and ≤ 1 day gap between each hospitalization record for general use purposes in HIV research. Future work may examine these definitions in other settings and populations.
format Article
id doaj-art-79e49a35b5774159aa14fca4f985f177
institution Kabale University
issn 1471-2288
language English
publishDate 2025-07-01
publisher BMC
record_format Article
series BMC Medical Research Methodology
spelling doaj-art-79e49a35b5774159aa14fca4f985f1772025-08-20T03:45:30ZengBMCBMC Medical Research Methodology1471-22882025-07-0125111010.1186/s12874-025-02602-5Identifying hospitalization episodes of care among people with and without HIV in British Columbia, CanadaSD Emerson0T McLinden1P Sereda2J Trigg3AM Yonkman4KA Salters5S Au6KW Kooij7MO Budu8VD Lima9R Barrios10RS Hogg11British Columbia Centre for Excellence in HIV/AIDSFaculty of Health Sciences, Simon Fraser UniversityBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSBritish Columbia Centre for Excellence in HIV/AIDSAbstract Background Hospitalizations are a resource-intensive form of healthcare use, particularly for persons with chronic conditions such as HIV. In standardized Canadian hospitalization databases, it can be unclear whether a hospitalization record is an independent hospitalization, a planned interhospital transfer, or an unplanned readmission. Misclassifying hospitalization records can bias metrics (e.g., counting transfers as readmissions can inflate readmission counts) and hence yield incorrect results. We compared definitions for combining sequential, related hospitalization records to create hospitalization episodes of care (HEoC) within a cohort of persons with and without HIV (PWH; PWoH) in British Columbia (BC), Canada. Methods Acute care hospitalization records (April 1992 to March 2020) were sourced from the Discharge Abstract Database within the Comparative Outcomes And Service Utilization Trends (COAST) study, a BC data linkage that includes samples of PWH and PWoH. Guided by published approaches and data quality considerations, we compared eight HEoC definitions applied to PWH and PWoH. Definitions varied by the date gap between records (0 day [same-day] or ≤ 1 day), and transfer indication (none required, populated transfer fields, one-way matching of hospital transfer identifiers, or two-way matching of hospital transfer identifiers). Comparisons were primarily informed by the percentage of multi-record HEoCs (HEoCs involving multiple hospitalization records, including interhospital transfers), and feasibility given data quality. Results The sample included 56,455 hospitalization records from 10,826 PWH, and 973,430 hospitalization records from 299,053 PWoH. Across the eight HEoC definitions, the percentage of multi-record HEoCs varied from 2.8 to 6.0% among PWH and 3.6 to 5.5% among PWoH. Definitions yielding the highest percentage of multi-record HEoCs combined records without requiring a transfer indication; definitions yielding the lowest percentage of multi-record HEoCs required two-way agreement of hospital identifiers. Patterns were generally comparable among PWH and PWoH, and similar in sensitivity analyses. Conclusions Various approaches can be used to define HEoCs. We recommended a balanced HEoC definition – requiring at least one populated hospital identifier field (without requiring matching of hospital identifiers) and ≤ 1 day gap between each hospitalization record for general use purposes in HIV research. Future work may examine these definitions in other settings and populations.https://doi.org/10.1186/s12874-025-02602-5
spellingShingle SD Emerson
T McLinden
P Sereda
J Trigg
AM Yonkman
KA Salters
S Au
KW Kooij
MO Budu
VD Lima
R Barrios
RS Hogg
Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada
BMC Medical Research Methodology
title Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada
title_full Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada
title_fullStr Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada
title_full_unstemmed Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada
title_short Identifying hospitalization episodes of care among people with and without HIV in British Columbia, Canada
title_sort identifying hospitalization episodes of care among people with and without hiv in british columbia canada
url https://doi.org/10.1186/s12874-025-02602-5
work_keys_str_mv AT sdemerson identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT tmclinden identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT psereda identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT jtrigg identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT amyonkman identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT kasalters identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT sau identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT kwkooij identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT mobudu identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT vdlima identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT rbarrios identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada
AT rshogg identifyinghospitalizationepisodesofcareamongpeoplewithandwithouthivinbritishcolumbiacanada