1
|
Woodman RJ, Mangoni AA. A comprehensive review of machine learning algorithms and their application in geriatric medicine: present and future. Aging Clin Exp Res 2023; 35:2363-2397. [PMID: 37682491 PMCID: PMC10627901 DOI: 10.1007/s40520-023-02552-2] [Citation(s) in RCA: 12] [Impact Index Per Article: 12.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/13/2023] [Accepted: 08/24/2023] [Indexed: 09/09/2023]
Abstract
The increasing access to health data worldwide is driving a resurgence in machine learning research, including data-hungry deep learning algorithms. More computationally efficient algorithms now offer unique opportunities to enhance diagnosis, risk stratification, and individualised approaches to patient management. Such opportunities are particularly relevant for the management of older patients, a group that is characterised by complex multimorbidity patterns and significant interindividual variability in homeostatic capacity, organ function, and response to treatment. Clinical tools that utilise machine learning algorithms to determine the optimal choice of treatment are slowly gaining the necessary approval from governing bodies and being implemented into healthcare, with significant implications for virtually all medical disciplines during the next phase of digital medicine. Beyond obtaining regulatory approval, a crucial element in implementing these tools is the trust and support of the people that use them. In this context, an increased understanding by clinicians of artificial intelligence and machine learning algorithms provides an appreciation of the possible benefits, risks, and uncertainties, and improves the chances for successful adoption. This review provides a broad taxonomy of machine learning algorithms, followed by a more detailed description of each algorithm class, their purpose and capabilities, and examples of their applications, particularly in geriatric medicine. Additional focus is given on the clinical implications and challenges involved in relying on devices with reduced interpretability and the progress made in counteracting the latter via the development of explainable machine learning.
Collapse
Affiliation(s)
- Richard J Woodman
- Centre of Epidemiology and Biostatistics, College of Medicine and Public Health, Flinders University, GPO Box 2100, Adelaide, SA, 5001, Australia.
| | - Arduino A Mangoni
- Discipline of Clinical Pharmacology, College of Medicine and Public Health, Flinders University, Adelaide, SA, Australia
- Department of Clinical Pharmacology, Flinders Medical Centre, Southern Adelaide Local Health Network, Adelaide, SA, Australia
| |
Collapse
|
2
|
Zhu M, Wei C, Yang X, Huang Y, Xu Y, Xiong Z. Lower haemoglobin-to-red blood cell distribution width ratio is independently associated with frailty in community-dwelling older adults: a cross-sectional study. BMJ Open 2023; 13:e069141. [PMID: 37423632 DOI: 10.1136/bmjopen-2022-069141] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 07/11/2023] Open
Abstract
OBJECTIVES The importance of blood cell markers in frailty has been studied. However, research on haemoglobin-to-red blood cell distribution width ratio (HRR) and frailty in older persons is still limited. We investigated the association between HRR and frailty in older adults. DESIGN Cross-sectional population-based study. SETTING Community-dwelling older adults older than 65 years were recruited from September 2021 to December 2021. PARTICIPANTS A total of 1296 community-dwelling older adults (age ≥65 years) in Wuhan were included in the study. MAIN OUTCOME MEASURES The main outcome was the presence of frailty. The Fried Frailty Phenotype Scale was used to evaluate the frailty status of the participants. Multivariable logistic regression analysis was performed to determine the relationship between HRR and frailty. RESULTS A total of 1296 (564 men) older adults were included in this cross-sectional study. Their mean age was 70.89±4.85 years. Receiver operating characteristic curve analysis showed that HRR is a good predictor of frailty in older people, the area under the curve (AUC) was 0.802 (95% CI: 0.755 to 0.849), and the highest sensitivity was 84.5% and the specificity was 61.9% with the optimal critical values 9.97 (p<0.001). Multiple logistic regression analysis indicated that lower HRR (<9.97) (OR: 3.419, 1.679 to 6.964, p=0.001) is independently associated with frailty in older people, even after adjusting confounding factors. CONCLUSION Lower HRR is closely associated with an increased risk of frailty in older people. Lower HRR may be an independent risk factor for frailty in community-dwelling older adults.
Collapse
Affiliation(s)
- Mengpei Zhu
- Institute of Geriatric Medicine, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
- Division of Gastroenterology, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
| | - Chao Wei
- Wuhan Geriatric Hospital, Wuhan, Hubei, China
| | - Xiongjun Yang
- Division of Gastroenterology, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
| | - Yumei Huang
- Division of Gastroenterology, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
| | - Yushuang Xu
- Division of Gastroenterology, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
| | - Zhifan Xiong
- Division of Gastroenterology, Liyuan Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China
| |
Collapse
|
3
|
McBurney MI, Tintle NL, Harris WS. Lower omega-3 status associated with higher erythrocyte distribution width and neutrophil-lymphocyte ratio in UK Biobank cohort. Prostaglandins Leukot Essent Fatty Acids 2023; 192:102567. [PMID: 36934703 DOI: 10.1016/j.plefa.2023.102567] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/09/2022] [Revised: 03/13/2023] [Accepted: 03/14/2023] [Indexed: 03/17/2023]
Abstract
High red blood distribution width (RDW) is associated with decreased red blood cell deformability, and high neutrophil-lymphocyte ratio (NLR) is a biomarker of systemic inflammation and innate-adaptive immune system imbalance. Both RDW and NLR are predictors of chronic disease risk and mortality. Omega-3 index (O3I) values have previously been shown to be inversely associated with RDW and NLR levels. Our objective was to determine if total plasma long chain omega-3 fatty acids (Omega3%) measured in the UK Biobank cohort were associated with RDW and NLR values. RDW- and NLR- relationships with Omega3% were characterized in 109,191 adults (58.4% female). RDW- and NLR-Omega3% relationships were inversely associated with Omega3% (both p < 0.0001). These cross-sectional associations confirm previous findings that increasing RDW and NLR values are associated with low O3I. The hypothesis that RDW and/or NLR values can be reduced in individuals with less-than optimal long chain omega 3 values need to be tested in randomized controlled intervention trials using EPA and/or DHA.
Collapse
Affiliation(s)
- Michael I McBurney
- Fatty Acid Research Institute, Sioux Falls, SD 57106, USA (MIM, NLT, WSH); Department of Human Health and Nutritional Sciences, University of Guelph, Guelph, ON N1G 2W1, Canada (MIM); Division of Biochemical and Molecular Biology, Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA 02111, USA (MIM).
| | - Nathan L Tintle
- Fatty Acid Research Institute, Sioux Falls, SD 57106, USA (MIM, NLT, WSH); Department of Population Health Nursing Science, College of Nursing, University of Illinois - Chicago, Chicago, IL 60612, USA (NLT)
| | - William S Harris
- Fatty Acid Research Institute, Sioux Falls, SD 57106, USA (MIM, NLT, WSH); Sanford School of Medicine, University of South Dakota, Sioux Falls, SD 57105, USA (WSH)
| |
Collapse
|
4
|
Alakare J, Kemp K, Strandberg T, Castrén M, Tolonen J, Harjola VP. Red cell distribution width and mortality in older patients with frailty in the emergency department. BMC Emerg Med 2023; 23:24. [PMID: 36894893 PMCID: PMC9998144 DOI: 10.1186/s12873-023-00801-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/10/2022] [Accepted: 03/01/2023] [Indexed: 03/11/2023] Open
Abstract
BACKGROUND The red cell distribution width (RDW) reflects the degree of heterogeneity of red blood cells. Elevated RDW is associated both with frailty and with increased mortality in hospital-admitted patients. In this study we evaluate whether high RDW values are associated with mortality in older emergency department (ED) patients with frailty, and if the association is independent of the degree of frailty. METHODS We included ED patients with the following criteria: ≥ 75 years of age, Clinical Frailty Scale (CFS) score of 4 to 8, and RDW % measured within 48 h of ED admission. Patients were allocated to six classes by their RDW value: ≤ 13%, 14%, 15%, 16%, 17%, and ≥ 18%. The outcome was death within 30 days of ED admission. Crude and adjusted odds ratios (OR) with 95% confidence intervals (CI) for a one-class increase in RDW for 30-day mortality were calculated via binary logistic regression analysis. Age, gender and CFS score were considered as potential confounders. RESULTS A total of 1407 patients (61.2% female), were included. The median age was 85 with an inter-quartile range (IQR) of 80-89, median CFS score 6 (IQR: 5-7), and median RDW 14 (IQR: 13-16). Of the included patients, 71.9% were admitted to hospital wards. A total of 85 patients (6.0%) died during the 30-day follow-up. Mortality rate was associated with RDW increase (p for trend < .001). Crude OR for a one-class increase in RDW for 30-day mortality was 1.32 (95% CI: 1.17-1.50, p < .001). When adjusted for age, gender and CFS-score, OR of mortality for one-class RDW increase was still 1.32 (95% CI: 1.16-1.50, p < .001). CONCLUSION Higher RDW values had a significant association with increased 30-day mortality risk in frail older adults in the ED, and this risk was independent of degree of frailty. RDW is a readily available biomarker for most ED patients. It might be beneficial to include it in risk stratification of older frail ED patients to identify those who could benefit from further diagnostic assessment, targeted interventions, and care planning.
Collapse
Affiliation(s)
- Janne Alakare
- Department of Emergency Medicine and Services, Helsinki University Hospital and University of Helsinki, Helsinki, Finland. .,Department of Geriatric Acute Care, Espoo Hospital, 2550 02070, City of Espoo, PL, Finland.
| | - Kirsi Kemp
- Department of Emergency Medicine and Services, Helsinki University Hospital and University of Helsinki, Helsinki, Finland
| | - Timo Strandberg
- University of Helsinki, Clinicum, and Helsinki University Hospital, Helsinki, Finland.,University of Oulu, Center for Life Course Health Research, Oulu, Finland
| | - Maaret Castrén
- Department of Emergency Medicine and Services, Helsinki University Hospital and University of Helsinki, Helsinki, Finland
| | - Jukka Tolonen
- Department of Internal Medicine, Helsinki University Hospital and University of Helsinki, Helsinki, Finland
| | - Veli-Pekka Harjola
- Department of Emergency Medicine and Services, Helsinki University Hospital and University of Helsinki, Helsinki, Finland
| |
Collapse
|
5
|
Kim DY, Hong DY, Kim SY, Park JJ, Kim JW, Park SO, Lee KR, Baek KJ. Prognostic value of red blood cell distribution width in predicting 3-month functional outcome of patients undergoing thrombolysis treatment for acute ischemic stroke. Medicine (Baltimore) 2021; 100:e27255. [PMID: 34664873 PMCID: PMC8447982 DOI: 10.1097/md.0000000000027255] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 01/12/2021] [Accepted: 08/30/2021] [Indexed: 01/25/2023] Open
Abstract
This study was performed to determine whether red blood cell distribution width (RDW) is associated with 3-month poor functional outcome in patients undergoing thrombolytic therapy for acute ischemic stroke.RDW was measured in patients with thrombolytic therapy in emergency department. Functional outcome was assessed after 3 months and poor functional outcome was defined as modified Rankin scale 3 to 6.A total of 240 patients were enrolled, and 82 (34.2%) had a poor functional outcome. The median RDW was significantly elevated in patients with a poor functional outcome compare with those with a good outcome. RDW was independently associated with a 3-month poor functional outcome (odds ratio 3.369, 95% confidence interval 2.214-5.125). The optimal RDW cutoff for predicting 3-month poor functional outcome was 12.8%, and the area under the curve for RDW was 0.818 (95% confidence interval 0.761-0.876). The area under the curve for RDW was higher in male patients than in female patients. The RDW correlated positively with the modified Rankin scale score after 3 months and the initial National Institutes of Health Stroke Scale score.Initial higher RDW level is related to a 3-month poor functional outcome in patients undergoing thrombolytic therapy for acute ischemic stroke.
Collapse
Affiliation(s)
- Dae Yong Kim
- Department of Emergency Medicine, Konkuk University Medical Center, Seoul, Republic of Korea
| | - Dae Young Hong
- Department of Emergency Medicine, Konkuk University School of Medicine, Seoul, Republic of Korea
| | - Sin Young Kim
- Department of Emergency Medicine, Konkuk University Medical Center, Seoul, Republic of Korea
| | - Jeong Jin Park
- Department of Neurology, Konkuk University School of Medicine, Seoul, Republic of Korea
| | - Jong Won Kim
- Department of Emergency Medicine, Konkuk University School of Medicine, Seoul, Republic of Korea
| | - Sang O. Park
- Department of Emergency Medicine, Konkuk University School of Medicine, Seoul, Republic of Korea
| | - Kyeong Ryong Lee
- Department of Emergency Medicine, Konkuk University School of Medicine, Seoul, Republic of Korea
| | - Kwang Je Baek
- Department of Emergency Medicine, Konkuk University School of Medicine, Seoul, Republic of Korea
| |
Collapse
|