PREDOMOS study, impact of a social intervention program for socially isolated elderly cancer patients: update to the study protocol for a randomized controlled trial

To improve patient recruitment, which was difficult from the beginning of the study, the steering committee decided to modify the inclusion criteria. Tumor type restrictions were lifted, allowing patients with locally advanced or metastatic cancer from all organs, including malignant hemopathies, to enter the study. Only patients with acute leukemia will be excluded, in the hope of including patients with a lifespan of more than 6 months. Furthermore, the treatment conditions selected were expanded to include patients cared for in a second line of oncologic treatment. Due to the inclusion of patients with various cancers, we also specified allowed oncological treatments: new generation hormonotherapy, targeted therapy, or immunotherapy with or without, concomitant or not, radiotherapy.

Otherwise, to simplify the exclusion criteria, the steering committee added a new inclusion criterion to the protocol, “Mini Mental State Examination (MMSE) score strictly over 24,” to automatically exclude patients unable to complete the QoL questionnaire without help and patients with dementia. We also decided to delete the exclusion criterion “patient with psychiatric troubles” because in the absence of cognitive impairment these patients are potential candidates like the others.

The inclusion criteria now include the following: (1) patient aged 70 years or older; (2) socially isolated or at risk of social isolation (modified Medical Outcomes Study Social Support Survey (m-MOS-SS) under 80% and/or isolated patient (caregiver living more than 50 km from the patient) and/or primary caregiver to his/her spouse (for a spouse who has a limited autonomy, neurodegenerative disease, cancer, etc., implying regular medical care for at least 3 months); (3) Scale for detection of geriatric frailties in cancer patients over 70 years old (G8 ONCODAGE) score ≤ 14; (4) Eastern Cooperative Oncology Group Performance Status (ECOG-PS) ≤ 2; (5) Activities of daily living (ADLs) score ≥ 4; (6) MMSE > 24; (7) lifespan > 6 months; (8) with locally advanced or metastatic cancer, including malignant hemopathies (except acute leukemia); (9) in first or second line of oncologic treatment by chemotherapy, new generation hormonotherapy, targeted therapy, or immunotherapy with or without, concomitant or not, radiotherapy; (10) signed informed consent; and (11) affiliation to the French social security system.

The exclusion criteria include: (1) patients with a second cancer; (2) protected adults under guardianship or curatorship; and (3) patient to be immediately directed into a rehabilitation and recuperative care service to benefit from treatment.

Home medical care for older cancer patients can be a relevant response to the issue of increasing expenditures related to aging. Indeed, studies have shown that these approaches could be cost-effective; namely, they were effective on a set of clinical criteria at an acceptable cost to the community or the health care system [2‐5]. For example, following an initial cost of implementation and equipment, home care interventions are accompanied by avoided costs of medical care (e.g., non-programmed re-hospitalizations, visits, examinations), as well as costs related to the financial burden for the patient. Consequently, cost analyses and cost-effectiveness analyses are crucial for planning proper distribution of health care resources and social/medical interventions in elderly cancer patient populations.

The aim of the economic analysis in the present study is to evaluate both the health benefits and costs of the two strategies and to calculate the incremental cost-effectiveness ratio of the innovative program assessed compared to the standard practice. As recommended by the French Health institution (HAS) [6], the clinical endpoint will be the number of quality-adjusted life years (QALYs) saved. These QALYs are determined by multiplying life years gained by a value of the utility associated (corresponding to patient’s quality of life) during the period under consideration [7]. As the scores in Quality of Life Questionnaire-Cancer (QLQ)-C30 are not utility-based, these scores will be mapped to the EuroQoL Five-levels (EQ-5D) [8, 9]. The QLQ-C30 at T3 will be used to cover the first 3 months (T0 to T3) and the QLQ-C30 at T6 to cover the period between T3 and T6. The time horizon will cover the period during which patients will be exposed to the intervention (i.e., 6 months). This will allow researchers to observe and collect prospectively all health and economic outcomes during the exposed period between the two groups. Therefore, no discounting will be necessary.

The perspective will be that of the society, and the following type of resources will be included in the cost analysis:

All resources will be observed, collected though the case report forms (CRFs), and valorized over the period between baseline and the 6 months follow-up. Several sources of unit costs will be investigated for this economic study based on HAS recommendations. Production costs through diagnostic-related groups from the Technical Agency for hospital information (ATIH) will be used for inpatient costs (sources: National Health system and hospital databases). For ambulatory care and examinations, national tariffs are recommended (sources: National databases for medical and paramedical acts, National official list prices for drugs and registers of pharmaceutical specialties, National Table for Biology).

Differences in health and costs outcomes’ means will be tested by using the Student t test (if normally distributed) and differences between distributions tested with the nonparametric Mann-Whitney U test (if means are not normally distributed). Data will be analyzed according to the intention-to-treat principle. Statistical methods will be investigated for dealing with missing data, and their potential impact (according to assumptions made to cope with missing data) on the CUA findings will be addressed in the sensitivity analysis. To address variability and to assess the generalizability of our results, univariate sensitivity analyses will be performed by varying health and cost variables one by one (including but not limited to incidence of sepsis, recurrence rates, or unit costs). The statistical uncertainty surrounding the incremental cost-utility ratio (due to sampling fluctuations for example) will be captured by a multivariate sensitivity analysis, using, among others, simulation methods such as non-parametric bootstrap methods [10]. In addition, cost-effectiveness acceptability curves will be constructed to represent decision uncertainty surrounding cost-effectiveness estimates [11].

At the time of manuscript submission, the status of the trial is “recruiting.”