Abstract
Background
The Multidimensional Evaluation of Elderly Person Form (SVaMA) is the official Comprehensive Geriatric Assessment (CGA) tool in most Italian regions for assessing medical, social, and functional needs of community-dwelling older people and developing an integrated care plan to meet them. The Multidimensional Prognostic Index (MPI) is a validated CGA-based tool for predicting mortality, other negative health outcomes and assessing multidimensional frailty. SVaMA’s wide diffusion in clinical practice prompted the development of an MPI version based on its data (MPI-SVaMA).
Aims
Assessing the usefulness in predicting mortality of a new digital version of the MPI-SVaMA (MPI-SVaMA Digit) through a simpler model.
Methods
In this retrospective cohort study, 12.020 community-dwelling older people (≥ 65 years) who underwent the SVaMA in Veneto, Italy were included. One-month and one-year mortality rates were retrieved from Registry Offices. The MPI-SVaMA Digit was obtained by assigning a risk category to each domain assessed in the SVaMA. Accuracy and precision were assessed using Area Under the Curve (AUC) and concordance index (C-index). The association between MPI-SVaMA Digit and mortality was evaluated through Cox regression analysis.
Results
The MPI-SVaMA Digit showed good accuracy and precision in predicting 1-month (AUC = 0.78; 95%CI 0.75–0.81, C-index = 0.78 95%CI 9.75–0.81) and 1-year mortality (AUC = 0.77; 95%CI 0.76–0.78, C-index = 0.72 95%CI 0.71–0.74). People at moderate and severe risk of multidimensional frailty showed, respectively a 4-fold and 12-fold increased mortality risk than the group at mild risk.
Conclusion
The new MPI-SVaMA Digit is an accurate prognostic tool for short- and long-term mortality useful to address clinical and organizational demands in community-dwelling older people.
Keywords: Multidimensional prognostic index, Community-dwelling, Older people, Gerontechnology, Comprehensive geriatric assessment
Introduction
The Comprehensive Geriatric Assessment (CGA) is the recognized gold standard method to assess medical, social, and functional needs of older people and has proved useful in developing an integrated and coordinated care plan to meet those needs [1]. The Multidimensional Evaluation of Elderly Person Form (SVaMA - Scheda di Valutazione Multidimensionale dell’Anziano) is the CGA official tool in most Italian Regions [2].
Initially introduced in Veneto Region in 2000 and then adopted by other nine different Italian Regions, the information provided by the SVaMA are typically used to develop personalized health-care interventions by the Public Local Health System, including access to homecare services and other public social-healthcare facilities, as well as admission to nursing homes and to community hospitals.
The Multidimensional Prognostic Index (MPI) is a CGA-based validated and widely used prognostic tool for assessing multidimensional frailty, originally developed with the aim of predicting short- and long-term mortality in hospitalized older people [3]. During the last 15 years, the MPI usefulness in predicting both short- and long-term mortality has been widely ascertained in several settings and in older persons affected by different diseases, making it one of the most reliable CGA-based tool for mortality prediction [4, 5].
Due to the wide diffusion of the SVaMA, twelve years ago, a version of the MPI based on information collected through the SVaMA was developed and validated. This MPI-SVaMA version has proven to be accurate in predicting both one-month and one-year mortality in community-dwelling older persons who required interventions by the Local Public Health System such as homecare assistance and nursing-home admission [6].
Thereafter the MPI-SVaMA was successfully used for clinical decision-making by testing the association of specific pharmacological treatments and mortality in relation to the severity grade of multidimensional frailty as assessed by the MPI-SVaMA. For example, in older patients with non-valvular atrial fibrillation, the stratification of multidimensional impairment through the MPI-SVaMA demonstrated that the treatment with anticoagulants, such as warfarin, was associated with reduced two-year mortality regardless of the grade of multidimensional frailty [7]. Similarly, statin treatment was linked to reduced three-year mortality in older patients with both coronary artery disease [8] and diabetes mellitus [9] irrespective of multidimensional frailty levels. Conversely, in a large population of 6818 older patients affected by dementia, the prescriptions of antidementia drugs (anticholinesterase and/or memantine) were associated with reduced two-year mortality [10] only in mild or moderate frail subjects (classes 1 and 2 of the MPI-SVaMA).
Although MPI-SVaMA appeared reliable and potentially very useful for clinical decisions in older people, its algorithm requires weighting information based on the specific population assessed, which relies on the data previously collected from the SVaMA, making the deployment of MPI-SVaMA quite complex in clinical practice.
For this reason, we tested the clinical accuracy of a digital version, population-independent MPI-SVaMA computation model (MPI-SVaMA Digit), in predicting one-month and one-year mortality in community-dwelling older people.
Methods
This study was a retrospective cohort study on consecutively enrolled community-dwelling older people aged 65 years or older, requiring public health care intervention or support by the Public Local Health Care System such as access to nursing homes, long-term care facilities, and home-care services.
Subjects included were evaluated at home or at the Local Public Health Geriatric Ambulatory from January 2004 to December 2010. Inclusion criteria were age over 65 years, capacity to give informed consent, and the availability of an assessed complete SVaMA. Participants were followed for 2 years through cities’ Registry Offices, and dates of deaths were retrieved from death certificates.
The SVaMA
The SVaMA is the standardized assessment tool officially recommended by the National Health Care System for conducting multidimensional evaluations of older adults living in the community or in nursing homes. When a resident requires assistance from the Public Local Health System, such as home-care services or institutionalization, a dedicated team including a general practitioner (GP), a nurse, a social worker, and a geriatrician acting as a consultant conduct a CGA to develop a personalized care plan. By using the SVaMA, the team gathers data on various aspects, including health, functional and cognitive impairments, social circumstances, and economic status. The SVaMA has been validated for its reliability, accuracy, and calibration. Currently, it is the recommended multidimensional assessment tool in many Italian regions, including Veneto, Trentino, Abruzzo, Basilicata, Calabria, Campania, Molise, Puglia, Sardegna, Sicilia, and Valle D’Aosta, collectively covering a population of approximately 21.7 million people. In the Supplementary Material, we reported the tool translated in English [2].
The SVaMa includes the assessment of nine domains as follows: 1) nursing care needs (VIP) through an 11-item scale evaluating the presence of (a) insulin-dependent diabetes mellitus, (b) heart failure, (c) liver failure with ascite, (d) tracheostomy, (e) oxygen therapy, (f) feeding gastrostomy and/or nasogastric tube, (g) central venous catheter, (h) bladder catheter, (i) ureterostomy, (j) nephrostomy, (k) pressure sores at the limbs or other sites; 2) the pressure sore risk (VPIA) using the Exton-Smith Scale; 3) the residual potential in terms of autonomy and recovery capacity (VPOT); 4) the total nursing and rehabilitative care needs (VSAN); 5) the three most impacting comorbidities using International Classification of Primary Care (ICPC); 6) social support (VSOC) relying on 16 items exploring the availability of a support network to help in managing the older person, house chores, and day and night supervision; 7) cognitive status (VCOG) using the Short Portable Mental Status Questionnaire (SPMSQ) and the evaluation of behavioural problems presence; 8) functional domain as assessed through the activities of daily living (VADL); 9) mobility (VMOB) through the Barthel Index assessing the independence in daily living activities, and autonomous movements.
The MPI-SVaMA digit
The MPI-SVaMA Digit is computed from nine SVaMA domains: cognitive status (VCOG), mobility (VMOB), activities of daily living (VADL), healthcare needs (VSAN), pressure sore risk (VPIA), residual skills (VPOT), nursing care needs (VIP), social support (VSOC), and main diagnosis, computed based on the SVaMA collected data.
As for all the different versions of the MPI, the MPI-SVaMA Digit was obtained by summing the risk category (score = 0, 0.5, or 1) associated with each domain’s score and then dividing the total by the number of nine domains explored. The resulting total MPI-SVaMA Digit score can be classified as mild (MPI-SVaMA Digit ≤ 0.33), moderate (0.33 < MPI-SVaMA Digit ≤ 0.66), or severe (MPI-SVaMA Digit > 0.66).
In the original version, which used the exact same dataset, the MPI-SVaMA was calculated using a weighted sum of each individual domain, then divided by the range of the included population [6]: a RECursive Partition and AMalgamation (RECPAM) algorithm was then used to identify the two cut-offs to divide the sample size in three categories. Further details have been previously reported elsewhere [6]. Even if this method led to a high precision and accuracy, this MPI-SVaMA had some limitations including the lack of the digital automatic collection of data derived from the SVaMA during the computation process and that it was not applicable to populations with different characteristics [6].The MPI-SVaMA Digit is similarly obtained by aggregating risk scores for the nine domains and using cut-off values for the interpretation of risk as in the standard MPI index, as reported in Table 1. Since both the MPI-SVaMA and the MPI-SVaMA Digit were computed on the same dataset but relying on two different models, no formal direct statistical comparison was carried out.
Table 1.
Domains included for the computation of the MPI-SVaMA digit
| Domains | SVaMA category | Mild risk = 0 | Moderate risk = 0.5 | Severe risk = 1 |
|---|---|---|---|---|
| Cognitive status |
VCOG (SPMSQ) |
0–3 | 4–7 | 8–10 |
| Mobility |
VMOB (Barthel Index) |
0–14 | 15–29 | 30–40 |
| Activities of daily living |
VADL (Barthel ADL) |
0–14 | 15–49 | 50–60 |
| Healthcare needs |
VSAN (nursing assistance SVaMA scale) |
0–5 | 6–20 | > 20 |
| Pressure sore risk |
VPIA (Exton-Smith Scale) |
0 | 1–10 | > 10 |
| Residual skills | VPOT | 20–25 | 5 | 0 |
| Nursing care need | VIP | 0 | 1–10 | > 11 |
| Social support | VSOC | 0–80 | 85–160 | 165–240 |
| Comorbidities | Main diagnosis | Other main medical conditions | Organ failure | Cancer |
VCOG, cognitive evaluation according to Short Portable Mental Status Questionnaire (SPMSQ); VMOB, mobility evaluation according to Barthel Index mobility Scale; VADL, functional assessment according to the Activities of Daily Living scale; VSAN healthcare needs evaluation according to nursing assistance SVaMA scale; VPIA, assessment of pressure sore risk according to the Exton-Smith Scale; VPOT, residual skills assessment; VIP, nursing care needs evaluation; VSOC, social network support evaluation
The MPI-SVaMA Digit software is freely available for download, along with the other MPI versions, at https://multiplat-age.it/index.php/en/tools and also on Play Store (https://play.google.com/store/apps/details?id=com.galliera.app&pli=1) and Apple Store (https://apps.apple.com/us/app/mpi-mobile/id6744017541) in the form of a mobile application.
Statistical analysis
Patients’ baseline characteristics were reported as means ± standard deviations (SD) or frequencies and percentages for continuous and categorical variables, respectively. Comparisons between genders were carried out using t-test or chi-squared.
The discriminatory power of the MPI-SVaMA Digit indices for 1-month and 1-year mortality prediction was assessed by displaying the Receiver Operating Characteristic (ROC) curve, and estimating the Area under the Curve (AUC) along with their 95% confidence intervals (95%CIs). The predictive performance of the model was also assessed by computing the Concordance Index (C-index), which reflects its ability to correctly rank patients based on their risk level and the timing of events. The association between MPI-SVaMA Digit and mortality, during one year of follow-up, was assessed using a Cox’s regression analysis, adjusted for age and sex. Taking those with an MPI-SVaMA Digit less than 0.33, representing robust older people, as reference, we reported the data for the other two categories as hazard ratios (HRs) with the correspondent 95%CI. To test the robustness of our results, we performed an interaction analysis sex by MPI-SVaMA Digit score and sex-stratified hazard ratios.
All statistical analyses were performed using SPSS 26.0, save for the C-index computation, run on STATA 18.5.
Results
Study population
Table 2 shows the descriptive statistics of the clinical and multidimensional characteristics of the participants. The mean age of the sample was 81.8 years (standard deviation 8 years), with females older than males (83.2 Vs. 79.5 years, respectively, p <.001).
Table 2.
Clinical and multidimensional characteristics of the 12,020 community-dwelling older participants included in the study who underwent a CGA-based SvAMA assessment
| Variable | All | Females | Males | p-value |
|---|---|---|---|---|
| Patients, n (%) | 12,020 | 7605 (63.3) | 4415 (36.7) | - |
| Age (mean ± standard deviation) | 81.8 ± 8 | 83.2 (7.7) | 79.5 ± 7.9 | < 0.001 |
| Main diagnosis, n (%) | < 0.001 | |||
| Dementia, n (%) | 3709 (30.9) | 2684 (35.3) | 1025 (23.2) | |
| Cancer, n (%) | 2834 (23.6) | 1288 (16.9) | 1546 (35.0) | |
| Hypokinetic syndrome, n (%) | 1340 (11.1) | 970 (12.8) | 370 (8.4) | |
| Cardiovascular disease, n (%) | 920 (7.7) | 623 (8.2) | 297 (6.7) | |
| Neurological disease, n (%) | 836 (7) | 528 (6.9) | 308 (6.9) | |
| Stroke, n (%) | 632 (5.3) | 388 (5.1) | 244 (5.5) | |
| Respiratory disease, n (%) | 294 (2.4) | 154 (2) | 140 (3.2) | |
| Hip fractures, n (%) | 290 (2.4) | 231 (3) | 59 (1.3) | |
| Other diseases, n (%) | 1165 (9.7) | 739 (9.7) | 426 (9.6) | |
| VCOG (mean ± standard deviation) | 5.34 ± 3.7 | 5.7 (3.6) | 4.8 (3.8) | < 0.001 |
| VMOB (mean ± standard deviation) | 28.2 ± 13.3 | 28.4 (13.2) | 27.9 (13.5) | 0.124 |
| VADL (mean ± standard deviation) | 40.1 ± 19.5 | 40.3 (19.5) | 39.8 (19.5) | 0.170 |
| VSAN (mean ± standard deviation) | 10.4 ± 11.4 | 9.5 (11.1) | 11.8 (11.8) | < 0.001 |
| VPIA (mean ± standard deviation) | 4.4 ± 6.1 | 4.4 (6) | 4.5 (6.1) | 0.111 |
| VPOT (mean ± standard deviation) | 0.42 ± 2.2 | 0.40 (2.1) | 0.48 (2.3) | 0.032 |
| VIP (mean ± standard deviation) | 5.5 ± 7.5 | 4.7 (7.1) | 6.8 (8.1) | < 0.001 |
| VSOC (mean ± standard deviation) | 149.1 ± 67.9 | 152.8 (68.1) | 142.6 (67) | < 0.001 |
| MPI-SVaMA Digit (mean ± standard deviation) | 0.46 ± 0.17 | 0.46 (0.17) | 0.48 (0.17) | < 0.001 |
| Deaths (1 month), n (%) | 201 (1.7) | 89 (1.2) | 112 (2.5) | < 0.001 |
| Deaths (12 months), n (%) | 1945 (16.2) | 913 (12) | 1032 (23.4) | < 0.001 |
VCOG, cognitive evaluation according to Short Portable Mental Status Questionnaire (SPMSQ); VMOB, mobility evaluation according to Barthel Index mobility Scale; VADL, functional assessment according to the Activities of Daily Living scale; VSAN healthcare needs evaluation according to nursing assistance SVaMA scale; VPIA, assessment of pressure sore risk according to the Exton-Smith Scale; VPOT, residual skills assessment; VIP, nursing care needs evaluation; VSOC, social network support evaluation
The final sample was composed of 12,020 participants, mostly females (n = 7605, 63.3%). The most common main diagnosis in the total sample (n = 3709, 30.9%) and in females (n = 2684, 35.3%) was dementia, whilst in males, it was cancer (n = 1546, 35%). If we take the tripartite subdivision used in MPI-SVaMA Digit into consideration, 23.6% were oncological patients, 11.7% reported organ failures and 64.7% was affected by other medical conditions. The distribution of main diagnosis was significantly different between men and women. The overall mortality incidence rates over one month and one year were, respectively, 1.7% (201 people) and 16.2% (1945 people).
Figure 1 shows the ROC curves of the MPI-SVaMA Digit in predicting mortality after 1 month and 1 year. The accuracy, as determined by the AUC, after one month was 0.78 (95% CI 0.75–0.81) whilst after 1 year was 0.77 (95% CI 0.76–0.78).
Fig. 1.
ROC Curve of the MPI-SVaMA Digit in predicting overall mortality after one month (on the left) and after one year (on the right)
The C-index of MPI-SVaMA Digit was 0.78 (95% CI 0.75–0.81) for 1-month mortality risk prediction and 0.72 (95% CI 0.71–0.74) for 1-year mortality risk prediction.
Table 3 shows the hazard ratios of death in the three MPI-SVaMA Digit categories of risk. People in the MPI-SVaMA Digit class 2 and 3 have, respectively, a 4-fold (HR = 4.079; 95% CI 3.278–5.076), and 12-fold higher mortality risk (HR = 12.29; 95% CI 9.836–15.346) than subjects included in the MPI-SVaMA Digit class 1.
Table 3.
Association between MPI-SVaMA digit and mortality in participants subdivided according to class of risk (MPI-SVaMA digit class 1, class 2 and class 3)
| HR | 95% CI per HR | |||
|---|---|---|---|---|
| Low | High | p-value | ||
|
MPI-SVaMA Digit Class 1 (MPI ≤ 0.33) (n = 2220) |
1 (reference) | - | - | - |
|
MPI-SVaMA Digit Class 2 (0.33 < MPI ≤ 0.66) (n = 7738) |
4.079 | 3.278 | 5.076 | < 0.001 |
|
MPI-SVaMA Digit Class 3 (MPI > 0.66) (n = 2062) |
12.286 | 9.836 | 15.346 | < 0.001 |
The data are reported as hazard ratios (HR), adjusted for age and gender
The same trend can be visually observed in the survival curves of Fig. 2, showing a clear separation among the curves of the different classes of risk of MPI-SVaMA Digit, i.e. participants included in the risk class 3 having a significantly lower survival rate than participants included in the MPI-SVaMA Digit class 2 and 1 class.
Fig. 2.
Survival curves for MPI-SVaMA Digit categories
Given the differences in descriptive statistics between males and females, an interaction analysis sex by MPI-SVaMA Digit score was performed showing a statistically significant effect of sex on the MPI-SVaMA Digit score (exp(B) = 0.293; p <.001).
Hazard ratios for mortality stratified by sex were, for males, HR = 4.500 (95% CI 2.868–7.060) in the MPI-SVaMA Digit class 2 and HR = 10.628 (95% CI 6.755–16.720) in the MPI-SVaMA Digit class 3, and for females, HR = 2.532 (95% CI 1.852–4.463) in the MPI-SVaMA Digit class 2 and HR = 4.995 (95% CI 3.628–6.878) in the MPI-SVaMA Digit class 3.
Discussion
The MPI-SVaMA Digit, computed through specifically developed software, with a new, simpler, and population-independent model, was found to be useful in forecasting short- and long-term mortality in a community-dwelling older population. The accuracy of this digital version of the MPI-SVaMA in predicting both short- and long-term mortality can be classified as good [5] (AUC between 0.7 and 0.8), being 0.78 at 1-month and 0.77 at 1-year. Notably, the good prognostic value of the instrument remains stable throughout time and comparable between one month and one year follow-up (0.78 Vs. 0.77, respectively). These findings were confirmed by the obtained C-indexes, that fall into the same score range, reflecting the model’s potential clinical usefulness in stratifying individuals based on their likelihood of survival over time.
A recent meta-analysis [5] showed the noteworthy accuracy of the MPI in predicting short- and long-term outcomes in different settings, especially in medical wards, as proven by several studies. In fact, MPI demonstrates a good accuracy (AUC = 0.79; 95% CI 0.76–0.81) and a very good precision (C-index 0.82; 95% CI 0.78–0.85) in predicting short-term mortality, maintaining a good accuracy after 6 months (AUC = 0.74; 95%CI: 0.71–0.76) and one year (AUC = 0.72; 95%CI: 0.69–0.76) [5].
In community settings, some relevant results on the prognostic efficacy of the MPI are available in several different study populations [11–13]. For example, in a study carried out in Sweden [14], using data from the Swedish National Study on Aging and Care in Kungsholmen (SNAC-K), the effectiveness of MPI has been verified in a population-based cohort analysis on 1331 older people residing at home or in institutions, followed for up to 13 years. In this study Time to Death in Years ranged from 3.8 to 9.0 based on MPI risk stratification and classes of age (72–78; 81–87; 90–99 years).
Findings from the InCHIANTI [15] study performed in Italy on 1453 community-dwelling older people reported a very good MPI accuracy with a C-index of 86.1 (95% CI 84.2–88.1) remaining stable throughout a very long follow-up period from 3, 6, 9, 12, up to 15 years.
Compared to the weighted MPI-SVaMA, the difference in the short-term mortality prediction’s accuracy is clinically negligible (0.83 Vs. 0.78) whilst the long-term one is essentially equivalent (0.79 Vs. 0.77). Given its undeniable computational simplicity, the MPI-SVaMA Digit looks to be the best option when SVaMA scores are available.
Using the MPI-SVaMA Digit, clinicians could monitor changes over time in the multidimensional frailty profile in older people who underwent a SVaMA evaluation, exploiting the MPI index’s clinimetric properties. Different versions of the MPI have been developed over time, and all of them, including MPI-SVaMA Digit, produce the same index measuring multidimensional frailty, facilitating the comparison of data across different settings and MPI versions [16].
The MPI-SVaMA Digit, similarly to the other MPI-derived indexes, can be expressed as a continuous value but also as a degree of multidimensional frailty severity (mild, moderate, severe), and subsequently can offer insight into the risk of mortality. The hazard ratios display a remarkable capability of the MPI to discriminate different risk categories of mortality in both genders with the association being stronger in male. The MPI-SVaMA has already shown its prognostic and clinical decision-making efficacy, guiding the prescription of several pharmacological therapies [7–10]. The present research also suggests its possible usefulness in predicting mortality regardless of population-specific computation weights, thus adding a crucial use-case to the tool. In fact, a limitation of the application of the original MPI-SVaMA algorithm is the necessity to rely on population-specific weight-based computation. The new MPI-SVaMA Digit model aimed to map the values resulting from the SVaMA to standardised thresholds irrespective of a specific population, and rather applicable in any context.
Furthermore, the MPI index can be interpreted universally by physicians unfamiliar with SVaMA evaluation. Conversely clinicians who are used to administering the SVaMA do not require further specific training for the MPI-SVaMA Digit.
Previous studies showed that MPI could offer insight to clinicians by supporting their clinical decisions, both regarding pharmacological therapeutic approach [8–10] and in more invasive procedures [17–19]. Moreover, the CGA-based multidimensional approach could identify specific impairments in single domains, which can in turn be used to guide the planning of a care pathway and a personalized intervention.
For instance, a moderate risk (“0.5”) in cognitive status could be addressed with a follow-up programme, whilst a severe risk could require referral to a specialistic center for further evaluation [20]. Similarly, risk of isolation and/or loneliness could be addressed with a specific intervention by social services [20].
Several mortality predictive tools have been developed over the years, primarily focusing on comorbidity [21]. The MPI-SVaMA Digit presents some particular characteristics such as relying on the CGA approach involving a multidisciplinary team for evaluation, thus not excluding people not able to self-report information. Moreover, this tool is based on clinical and functional indexes developed and validated in older people. All these features allow MPI-SVaMA Digit to intercept a wider population of frail individuals compared with previously developed mortality prediction tools. The data of the present study confirm the reliability and accuracy of the CGA-based MPI tool for predicting mortality, thus adding a considerable feature to the SVaMA, which was originally developed only for the assessment of the needs for assistance and interventions, without burdening healthcare professionals and patients with a additional questionnaire. The literature showed that CGA could reduce [22] mortality rates as well as functional and cognitive impairments, therefore the MPI-SVaMA Digit, which is based on this approach, may improve the same outcomes.
The findings of this study must be interpreted within its limitations. First, the MPI-SVaMA Digit was calculated in a specific population, i.e., older people requiring a nursing home admission or homecare services: therefore, even if we included a large population, it was not representative of the overall community-dwelling older people. Future research should therefore focus on seeking further external validation, demonstrating consistent performance in different cohorts, hence strengthening its generalizability beyond the Veneto region population. Second, for some domains, such as comorbidities, we used a simplified classification of the diseases.
In conclusion, this new digital version of MPI-SVaMA showed similar accuracy and precision compared to the older version, but it is easier and quicker to apply by health-professionals in different settings and thus potentially useful in detangling clinical and organizational demands in older people.
Author contributions
Writing– original draft preparation: Morganti, W., Seminerio, E., Veronese, N.; Writing– review and editing: Pilotto, A., Corti, M.C., Maggi, S., Ferrucci, L.; Formal analyses and investigation: Veronese, N., Morganti, W., Seminerio, E.; Conceptualization: Pilotto, A., Maggi, S., Corti M.C., Funding acquisition: Pilotto, A; Data curation: Massone, C., Fielding, P.
Funding
The presented work is funded by the Project “Development and implementation of common strategy for the management of community-dwelling older subjects with multimorbidity and polypharmacy: integration with a multicomponent intervention platform by using domotic, robotic and telecare systems” (MULTIPLAT_AGE), NET-2016-02361805.
Data availability
The datasets generated and/or analysed during the present study are not publicly available. However, the datasets are available from the authors on reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Parker SG et al (2018) «What is Comprehensive Geriatric Assessment (CGA)? An umbrella review», Age and Ageing, vol. 47, fasc. 1, pp. 149–155, gen. 10.1093/ageing/afx166 [DOI] [PubMed]
- 2.«Form for the Multidimensional Evaluation of Elderly Patients [Online]. Disponibile su: https://web.archive.org/web/20151223183859/http://www.ulss16.padova.it/all/SchedaSvama_EN.pdf»
- 3.Pilotto A et al (2008) feb., «Development and Validation of a Multidimensional Prognostic Index for One-Year Mortality from Comprehensive Geriatric Assessment in Hospitalized Older Patients», Rejuvenation Research, vol. 11, fasc. 1, pp. 151–161. 10.1089/rej.2007.0569 [DOI] [PMC free article] [PubMed]
- 4.Zampino M et al (2022) «Biomarkers of aging in real life: three questions on aging and the comprehensive geriatric assessment», GeroScience, vol. 44, fasc. 6, pp. 2611–2622, dic. 10.1007/s11357-022-00613-4 [DOI] [PMC free article] [PubMed]
- 5.Veronese N et al (2024) «Clinical prognostic factors for older people: A systematic review and meta-analysis». Ageing Res Rev 98:102345. 10.1016/j.arr.2024.102345 [DOI] [PubMed] [Google Scholar]
- 6.Pilotto A et al (2013) apr., «Development and Validation of a Multidimensional Prognostic Index for Mortality Based on a Standardized Multidimensional Assessment Schedule (MPI-SVaMA) in Community-Dwelling Older Subjects», Journal of the American Medical Directors Association, vol. 14, fasc. 4, pp. 287–292. 10.1016/j.jamda.2013.01.005 [DOI] [PMC free article] [PubMed]
- 7.Pilotto A et al (2016) «Warfarin Treatment and All-Cause Mortality in Community‐Dwelling Older Adults with Atrial Fibrillation: A Retrospective Observational Study», J American Geriatrics Society, vol. 64, fasc. 7, pp. 1416–1424, lug. 10.1111/jgs.14221 [DOI] [PMC free article] [PubMed]
- 8.Pilotto A et al (2016) «Relation of Statin Use and Mortality in Community-Dwelling Frail Older Patients With Coronary Artery Disease», The American Journal of Cardiology, vol. 118, fasc. 11, pp. 1624–1630, dic. 10.1016/j.amjcard.2016.08.042 [DOI] [PMC free article] [PubMed]
- 9.Pilotto A et al (2015) «Statin Treatment and Mortality in Community-Dwelling Frail Older Patients with Diabetes Mellitus: A Retrospective Observational Study», PLoS ONE, vol. 10, fasc. 6, p. e0130946, giu. 10.1371/journal.pone.0130946 [DOI] [PMC free article] [PubMed]
- 10.Pilotto A et al (2018) feb., «Association of Antidementia Drugs and Mortality in Community-Dwelling Frail Older Patients With Dementia: The Role of Mortality Risk Assessment», Journal of the American Medical Directors Association, vol. 19, fasc. 2, pp. 162–168. 10.1016/j.jamda.2017.08.017 [DOI] [PMC free article] [PubMed]
- 11.Cella A et al (2021) nov., «Multidimensional Frailty Predicts Mortality Better than Physical Frailty in Community-Dwelling Older People: A Five-Year Longitudinal Cohort Study», IJERPH, vol. 18, fasc. 23, p. 12435. 10.3390/ijerph182312435 [DOI] [PMC free article] [PubMed]
- 12.Veronese N et al (2021) «Mortality attributable to COVID-19 in nursing home residents: a retrospective study», Aging Clin Exp Res, vol. 33, fasc. 6, pp. 1745–1751, giu. 10.1007/s40520-021-01855-6 [DOI] [PMC free article] [PubMed]
- 13.Veronese N et al (2022) «Multidimensional frailty and quality of life: data from the English Longitudinal Study of Ageing», Qual Life Res, vol. 31, fasc. 10, pp. 2985–2993, ott. 10.1007/s11136-022-03152-9 [DOI] [PMC free article] [PubMed]
- 14.Angleman SB, Santoni G, Pilotto A, Fratiglioni L (2015) A.-K. Welmer, e on behalf of the MPI_AGE Project Investigators, «Multidimensional Prognostic Index in Association with Future Mortality and Number of Hospital Days in a Population-Based Sample of Older Adults: Results of the EU Funded MPI_AGE Project», PLoS ONE, vol. 10, fasc. 7, p. e0133789, lug. 10.1371/journal.pone.0133789 [DOI] [PMC free article] [PubMed]
- 15.Pilotto A et al (2021) «Association Between the Multidimensional Prognostic Index and Mortality During 15 Years of Follow-up in the InCHIANTI Study», The Journals of Gerontology: Series A, vol. 76, fasc. 9, pp. 1678–1685, ago. 10.1093/gerona/glaa237 [DOI] [PMC free article] [PubMed]
- 16.Veronese N et al (2021) «Prevalence of multidimensional frailty and pre-frailty in older people in different settings: A systematic review and meta-analysis». Ageing Res Rev 72:101498. 10.1016/j.arr.2021.101498 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lai S et al (2020) «Association between Multidimensional Prognostic Index and Hospitalization and Mortality among Older Adults with Chronic Kidney Disease on Conservative or on Replacement Therapy», JCM, vol. 9, fasc. 12, p. 3965, dic. 10.3390/jcm9123965 [DOI] [PMC free article] [PubMed]
- 18.Custodero C et al (2021) «Multidimensional prognostic index (MPI) predicts non-invasive ventilation failure in older adults with acute respiratory failure». Arch Gerontol Geriatr 94:104327. 10.1016/j.archger.2020.104327 [DOI] [PubMed] [Google Scholar]
- 19.Veronese N et al (2020) «Enteral tube feeding and mortality in hospitalized older patients: A multicenter longitudinal study», Clinical Nutrition, vol. 39, fasc. 5, pp. 1608–1612, mag. 10.1016/j.clnu.2019.07.011 [DOI] [PubMed]
- 20.Pilotto A et al (2024) «A multicomponent personalized prevention program in the primary care setting: a randomized clinical trial in older people with noncommunicable chronic diseases (Primacare_P3 study)», Trials, vol. 25, fasc. 1, p. 611, set. 10.1186/s13063-024-08413-1 [DOI] [PMC free article] [PubMed]
- 21.Exmann CJC, Kooijmans ECM, Joling KJ, Burchell GL, Hoogendijk EO, Van Hout eHPJ (nov. 2024) «Mortality prediction models for community-dwelling older adults: A systematic review». Ageing Res Rev 101:102525. 10.1016/j.arr.2024.102525 [DOI] [PubMed]
- 22.Ellis G et al (2017) «Comprehensive geriatric assessment for older adults admitted to hospital», Cochrane Database of Systematic Reviews, vol. fasc. 9, set. 2017. 10.1002/14651858.CD006211.pub3 [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analysed during the present study are not publicly available. However, the datasets are available from the authors on reasonable request.