Keywords
Medical Informatics - Clinical Decision Support System - Implementation, Expert System
1. Introduction
Clinical decision support (CDS) systems are central to promoting safe and evidence-based
decisions in an increasingly complex healthcare ecosystem. This manuscript represents
the synopsis for the Decision Support section of the 2024 International Medical Informatics
Association (IMIA) Yearbook. This synopsis supplements the review paper on CDS related
to precision medicine, authored by Sulieman et al [[1] ref to DS survey YB24]. The aim of this synopsis is to summarize significant research
in the CDS domain and to select the best papers published in this field in 2023.
2. Methods
We used an updated search strategy that we co-developed for the CDS synopsis in the
2023 IMIA Yearbook [[2]]. We are including the actual query in [Table 1] to support future Yearbook contributors. Similar to the synopsis for 2022, we searched
for relevant papers in PubMed® using the revised search strategy. We included journal
articles published in 2023 in English with a series of keywords either in the MeSH
terms or in the paper's title or abstract. For the MeSH terms, we considered, “Clinical
Decision Support”, “Expert systems”, Medical Order Entry Systems”, “Decision Support
Techniques”, and “Decision Support Systems, Management”. For the abstracts and title
search, we used the terms, “Best Practice Alert”, “Clinical Decision Support”, “Decision
Support Techniques”, “Decision Support Systems”, “Expert systems”, “Medication Alert
Systems”, and “Computerized Provider Order Entry System”. The final query is shown
in [Table 1].
Table 1.
Search Strategy for CDS Literature Review.
|
2023[DP] NOT pubstatusaheadofprint AND
(journal article[PT]) AND
English[LA] AND
(
(Clinical Decision Support[MeSH Terms]) OR (Expert systems[Mesh]) OR (Medical Order
Entry Systems[Mesh]) OR (Decision Support Techniques[MeSH Terms]) or (Decision Support
Systems, Management [MeSH Terms]) OR
„Best Practice Alert“[title/abstract] OR
„Clinical Decision Support“[title/abstract] OR
„Decision Support Techniques“[title/abstract] OR
„Decision Support Systems“[title/abstract] OR
„Expert systems“[title/abstract] OR
„Medication Alert Systems“[title/abstract] OR
„Computerized Provider Order Entry System“ [title/abstract]
)
|
We reviewed the query results in three steps: title only review, followed by abstract
review, and then full text review as described below. We chose to first triage using
a title only review due to the large volume of results from the query (see Results section for more details). At this stage, reviewers were blinded to country of origin,
institution, or authors. We included any articles that were selected for further review
by at least one of section editors (CUL and VS). Following the title only review,
we proceeded to screen the remaining articles using their abstracts and again retained
papers that were selected by either reviewer. We then individually examined the full
text of all remaining articles and held a virtual discussion for each article that
was selected by both editors. Through consensus, we identified the candidate best
papers. The full IMIA Yearbook editorial committee determined the three finalists
and an honorable mention.
3. Results
The search query resulted in 1,948 unique references. After the title review, 116
references were included by one or both of the section editors (CUL and VS). After
screening the abstracts, 46 articles were identified with six selected by both editors.
After full text review, four best paper candidates were selected by consensus ([Table 2]). Of these, two papers were published in Applied Clinical Informatics and one each
in NPJ Microgravity and Nature Communication, respectively. We provide further analyses
of the four best paper candidates using the Donabedian's structure-process-outcomes
model [[3]]. After external review of best paper candidates, the IMIA Yearbook editorial committee
selected three candidates as the best papers ([Table 3]) with the most significant contributions in the field of clinical decision support
for 2023. The remaining paper is included in this synopsis as an honorable mention.
A summary of the best papers can be found in the appendix of this synopsis.
Table 2:
Best paper candidates for the Clinical Decision Support Section.
|
PMID
|
Title
|
First Author
|
Journal
|
DOI
|
|
38092360*
|
How Safe are Outpatient Electronic Health Records? An Evaluation of Medication-Related
Decision Support using the Ambulatory Electronic Health Record Evaluation Tool
|
Co Z
|
Appl Clin Inform
|
10.1055/s-0043-1777107
|
|
36792057*
|
Improving Pediatric Intensive Care Unit Discharge Timeliness of Infants with Bronchiolitis
Using Clinical Decision Support
|
Martin B
|
Appl Clin Inform
|
10.1055/a-2036-0337
|
|
37344482^
|
The value of a spaceflight clinical decision support system for earth-independent
medical operations
|
Russell BK
|
NPJ Microgravity
|
10.1038/s41526-023-00284-1
|
|
37198160*
|
A randomized clinical trial assessing the effect of automated medication-targeted
alerts on acute kidney injury outcomes
|
Wilson FP
|
Nat Commun
|
10.1038/s41467-023-38532-3
|
|
*Best paper finalist, ^Honorable Mention
|
Table 3.
Selection of best papers for the 2024 IMIA Yearbook of Medical Informatics for the
section Decision Support. The articles are listed in alphabetical order of the first
author's last name.
|
Section Decision Support
|
|
• Co Z, Classen DC, Cole JM, Seger DL, Madsen R, Davis T, McGaffigan P, Bates DW.
How Safe are Outpatient Electronic Health Records? An Evaluation of Medication-Related
Decision Support using the Ambulatory Electronic Health Record Evaluation Tool. Appl
Clin Inform. 2023 Oct;14(5):981-991. doi: 10.1055/s-0043-1777107.
• Martin B, Mulhern B, Majors M, Rolison E, McCombs T, Smith G, Fisher C, Diaz E,
Downen D, Brittan M. Improving Pediatric Intensive Care Unit Discharge Timeliness
of Infants with Bronchiolitis Using Clinical Decision Support. Appl Clin Inform. 2023
Mar;14(2):392-399. doi: 10.1055/a-2036-0337.
• Wilson FP, Yamamoto Y, Martin M, Coronel-Moreno C, Li F, Cheng C, Aklilu A, Ghazi
L, Greenberg JH, Latham S, Melchinger H, Mansour SG, Moledina DG, Parikh CR, Partridge
C, Testani JM, Ugwuowo U. A randomized clinical trial assessing the effect of automated
medication-targeted alerts on acute kidney injury outcomes. Nat Commun. 2023 May 17;14(1):2826.
doi: 10.1038/s41467-023-38532-3.
|
4. Discussion and Outlook
4. Discussion and Outlook
The four best paper candidates were analyzed using various structural, process, and
outcome characteristics. Structural aspects include the clinical setting where the
CDS was deployed or studied and the type of CDS. The clinical setting of interest
included primary care, pediatric intensive care unit (ICU), inpatient hospital setting,
and extraterrestrial (i.e., earth-independent) clinical operations during a spaceflight.
The type of CDS included a phone alert to notify clinicians about pediatric patients
that were ready for transfer from ICU to hospital floor [[4]]; medication related electronic alerts [[5]]; an active, interruptive alert with targeted information on labs (creatinine levels)
and medications related to kidney function [[6]]; spaceflight CDS scenarios for diagnostic support, treatment, monitoring, and providing
reference information [[7]]. Clinical conditions of interest include acute kidney injury, bronchiolitis, life-threatening
scenarios (e.g., events related to ABCDE – airway, breathing, circulation, disability, and exposure),
and hypertension.
In terms of process characteristics, we primarily analyzed the nature of CDS evaluation
(e.g., study design; multisite vs. single-site studies; provider-facing vs patient-facing CDS; early-stage vs. effectiveness
studies). The majority of the CDS systems were clinician-centered, with limited to
no involvement of patients or patient-centered factors such as shared-decision making
and social drivers of health. Martin et al. [[4]] conducted a single-site, early-stage, implementation study to evaluate a CDS for
identifying likely patient candidates that are ready for transfer from the ICU to
the hospital floor. Co et al. [[5]] conducted a cross-sectional, multi-center study to evaluate medication-related
CDS tools in ambulatory setting. The study by Wilson et al. [[6]] was a prospective, multi-center, open-label, randomized controlled trial to study
the effect of an alert-based CDS on rates of discontinuation or cessation of nephrotoxic
medications. The paper by Russel et al. [[7]] was a conceptual overview and demonstration of CDS systems that may be useful for
clinical operations in spaceflight missions. Notably, this work took a broader perspective
of the value and extent of CDS systems can offer in a resource-constrained and highly
controlled environment, while the other three studies were relatively narrow in terms
of scope and largely focused on evaluation of rule-based CDS. The outcomes of interest
in the four studies included inpatient mortality, progression of acute kidney injury,
potential benefits of CDS for a spaceflight mission with an autonomous crew performing
medical operations, rates and types of medication orders, length of stay in intensive
care units versus hospital, and hospitalization costs.
