Original article
Development of a new predictive model for polypathological patients. The PROFUND index

https://doi.org/10.1016/j.ejim.2010.11.012Get rights and content

Abstract

Background

There is a concern about the accuracy of the available prognostic indexes when applying them to the emergent population of polypathological patients (PP).

Methods

To develop a 1-year mortality predictive index on PP, we developed a multicenter prospective cohort-study recruiting 1.632 PP after hospital discharge, outpatient clinics, or home hospitalization, from 33 hospitals. Potential risk factors were obtained in the 1.525 PP who completed follow-up. Each factor independently associated with mortality in the derivation cohort (757 PP from western hospitals) was assigned a weight, and risk scores were calculated by adding the points of each factor. Accuracy was assessed in the validation cohort (768 PP from eastern hospitals) by risk quartiles calibration, and discrimination power, by ROC curves. Finally, accuracy of the index was compared with that of the Charlson index.

Results

Mortality in the derivation/validation cohorts was 35%/39.5%, respectively. Nine independent mortality predictors were identified to create the index (age ≥ 85 years, 3points; No caregiver or caregiver other than spouse, 2points; active neoplasia, 6points; dementia, 3points; III–IV functional class on NYHA and/or MRC, 3points; delirium during last hospital admission, 3points; hemoglobinemia < 10 g/dl, 3points; Barthel index < 60 points, 4points; ≥ 4 hospital admissions in last 12 months, 3points). Mortality in the derivation/validation cohorts was 12.1%/14.6% for patients with 0–2points; 21.5%/31.5% for those with 3–6 points; 45%/50% for those with 7–10 points; and 68%/61.3% for those with ≥ 11points, respectively. Calibration was good in derivation/validation cohorts, and discrimination power by area under the curve was 0.77/0.7. Calibration of the Charlson index was good, but discrimination power was suboptimal (area under the curve, 0.59).

Conclusions

This prognostic index provides an accurate and transportable method of stratifying 1-year death risk in PP.

Introduction

Polypathological patients (PP) have become an emerging population in most clinical arenas [1], [2], [3]. Their prevalence in Primary Care, as well as in different medical and surgical hospital areas, is notably high and will most likely increase in the forthcoming years [1], [4], [5]. The term PP is patient-centered and applies those patients suffering from chronic diseases from 2 or more of eight predefined categories; these categories were established by a panel of experts using criteria of end-effect on function of key organs (independent of the primary disease), frequent chronic conditions with high mortality/potential of becoming unstable, or frequent comorbidities when mental/functional impairment thresholds were definitively reached (Table 1) [1], [2], [3]. Their complexity, disease and symptom burden, clinical vulnerability, poor health-related quality of life, tendency towards functional deterioration, and the impact on relatives and caregivers have been well described [1], [6], [7]. Because of all these factors, their mortality during hospitalization and in Primary Care follow-up is outstandingly high [1], [2], [3], [6].

Survival prognostication is a cornerstone for clinicians in patient management and for health-care providers in designing health policies. It is not only a professional but also an ethical concern to clear possible areas of uncertainty in this issue that could avoid baseless nihilist attitudes or, on the contrary, diagnostic–therapeutic fury-futility. This is especially important in high-risk populations in order to reassess care goals; redefine medically necessary therapies; focus on symptom control; assess other physical, psychosocial, and spiritual problems‘ and consider earlier palliative care. With the knowledge of a reasonable precise prognosis, clinicians can feel more comfortable raising important issues like care goals, treatment preferences, advanced planning, and clinical therapeutic options with patients and their families [8], [9].

When assessing prognosis in PP populations, we are faced with many difficulties. Organ- or disease-specific indexes are not suitable due to the usual co-protagonism and similar weight of two or more chronic disabling diseases in PP. More generic and recent instruments require subjective assessments of risk by clinicians, include disused functional scales, use difficult algorithms, or are based on administrative data [9], [10], [11], [12]. Hence, they are not often used in routine practice, and most clinicians and investigators continue to use the Charlson–Deyo index as the gold-standard tool when referring to prognosis in patients with comorbidity [13], [14], [15]. Nevertheless, the Charlson–Deyo index has been around for over 20 years; during this time, new diagnostic–therapeutic options have drastically changed the course of many of the included conditions (like peptic ulcer, cardiovascular diseases, AIDS, or some neoplasias), so there is a generalized concern among clinicians that it may have lost some of its accuracy [15].

For all of the abovementioned reasons, we developed this multicenter project with the aim of obtaining an accurate prognostic tool specially designed for this vulnerable population, and then compare its fitness to the Charlson–Deyo index.

Section snippets

Patients and methods

This was an observational prospective, multi-institutional study carried out by researchers from the Polypathological Patient and Advanced Age Study Group of the Spanish Internal Medicine Society. The study inclusion period ranged from February 2007 to June 2008 (17 months).

Results

A total of 1632 PP (75% hospitalized, 17.5% outpatient, and 7.5% at-home patients) were included in the study, and 93.44% of them (N = 1525) completed the 12-month follow-up. The main demographical, clinical, and care features of the whole inclusion cohort have already been described [24]. Division of the cohort was performed with the patients who completed the follow-up. The compared main basal features of patients in the derivation (n = 757) and validation cohort (n = 768 patients) are detailed in

Discussion

We have developed and validated a new accurate and easy-to-perform prognostic index specifically designed for polypathological patient populations. The PROFUND index includes demographical (age), clinical (presence of neoplasia, dementia, disabling dyspnea, and delirium in last hospital admission), laboratory (hemoglobin), functional (BI), socio-familial (No caregiver or caregiver other than spouse), and care (number of hospitalizations in last 12 months) variables. The cut-off point of 85 years

Learning points

  • Polypathological patients conform a homogeneous population in our hospitals. They are characterized by their high 1-year death risk.

  • Nine risk factors were independently associated to mortality and used to develop the PROFUND prognostic index. These were age ≥ 85 years, 3 points; No caregiver or caregiver other than spouse, 2 points; active neoplasia, 6 points; dementia, 3 points; III–IV functional class on NYHA and/or MRC, 3 points; delirium during last hospital admission, 3 points;

Conflict of interest

The authors have no conflicts of interest to report.

Author's contribution

All authors have contributed substantially to the work.

Ethics committee approval

The present study has been approved by the ethics committee of Hospitales Universitarios Virgen del Rocío.

List of researchers from the PROFUND project

Máximo Bernabeu-Wittel1, Aurelio Fuertes-Martín2, Bosco Barón-Franco3, José Murcia-Zaragoza4, Carmen Ramos-Cantos5, Antonio Alemán6, Antonio Fernández-Moyano7, Manuel Ollero-Baturone1, Javier Galindo-Ocaña1, Alberto Escalera-Zalvide3, María Sanchez-Ledesma2, Antonia Mora-Rufete8, José Barbé Gil-Ortega9, Lourdes Moreno-Gaviño1, Virginia Rodríguez-Martínez10, Carolina Luque-Amado11, Manuel Romero-Jiménez12, María Maiz-Jiménez11, Mª Ángeles Soria-López13, Manuel Rincón-Gómez1, Manuel Bayón-Sayago1

Acknowledgment

The study was supported by Instituto de Salud Carlos III (ISCIII, FIS 07/0047), Ministerio de Sanidad y Consumo, Spain.

References (35)

  • World Health Organization

    Atlas of health in Europe 2nd Edition 2008

  • EUROSTAT

    Europe in figures. EUROSTAT Year Book 2009

  • Department of Health

    Supporting people with long term conditions. An NHS and Social Care Model to support local innovation and integration

  • W.A. Knaus et al.

    The SUPPORT prognostic model: objective estimates of survival for seriously ill hospitalized older adults

    Ann Intern Med

    (1995)
  • L.C. Walter et al.

    Development and validation of a prognostic index for 1-year mortality in older adults after hospitalization

    JAMA

    (2001)
  • J.M. Teno et al.

    Prediction of survival for older hospitalized patients; the HELP survival model

    J Am Geriatr Soc

    (2000)
  • D.A. Southern et al.

    Comparison of the Elixhauser and Charlson/Deyo methods of comorbidity measurement in administrative data

    Med Care

    (2004)
  • Cited by (124)

    View all citing articles on Scopus
    1

    On behalf of the PROFUND RESEARCHERS.

    View full text