First referral to an integrated onco-palliative care program: a retrospective analysis of its timing

Cochin Hospital is a tertiary care hospital treating around 4500 new cancer patients each year, with an oncology ward and three other medical specialty wards (gastroenterology, pneumology, dermatology) that have an oncologic activity of care.

At the time of the study, the oncology ward consists in a 9-beds inpatients unit and an 11-beds outpatient clinics. Medical staff is made up of three attending physicians and three fellow physicians, all advising two residents for inpatients and three residents for the outpatient clinic ambulatory patients.

The PCT consists in 2.5 full time equivalent physicians, all being palliative care specialists, 2.5 full time equivalent nurses and one secretary assistant.

Social workers and psychologists collaborate with both teams.

The IOPC program has been developed as a specific organization involving the PCT and the oncology staff.

This organization relies on weekly multidisciplinary onco-palliative meetings (OPM), which are attended by both the PCT and the oncology staff, i.e. physicians, head nurses, social workers and psychologists. Physicians of the PCT are in charge of moderating, keeping record of each meeting and reporting any decision and its rationale in the patient’s health record. The oncologists choose to refer patients to these meetings regarding the following criteria: situation of incurability and necessity to discuss goals and organization of care to anticipate the trajectory of care. Discussions take into account expected benefit of treatment on survival and quality of life, proportionality of care, and patient’s preferences. Decisions may be to pursue or change antitumoral therapies, associated or not with the introduction of the PCT, or to provide palliative care only. These decisions are then submitted and discussed with the patient. Later on, patients are followed-up by both the referent oncologist and the PCT, if deemed appropriate, in consultations, outpatient clinics, or inpatient acute care setting. For all patients discussed, goals and organization of care can be updated at following OPM, up to patient’s death. A part of OPM is also dedicated to deceased patients to review the trajectory of care, and the aggressiveness of care near the end-of-life.

Along with this organization, inpatients or outpatients can be referred to the PCT in an on-demand way, before being discussed at the OPM, if they are presenting with urgent needs (urgent psycho-physical symptoms, urgent need for shared-decision such as serious complications that require discussion on the appropriate intensity of care).

The first referral to IOPC is defined either as the first report at OPM or as the first referral to the PCT (Fig. 1).

This study included the historical cohort of patients first reported at OPM between January 1st, 2011 and December 31, 2013. Non-eligibility criteria were: unavailability of patient’s health records, patients presenting with non-oncologic disease or curative cancer and first referral to OPM after death. All included patients were then followed-up until death or until December 31, 2016.

We collected data from patients’ files concerning: 1) social and clinical characteristics of patients: age and gender, primary cancer site, date of initial diagnosis, date at which disease was deemed incurable (i.e. without curative treatment options due to metastatic stage or inoperability, or both), indicators of social vulnerability (precarious living conditions, living alone, in charge of some relative, spouse diagnosed with serious disease, incapability to express wills from somatic causes), other health risks (active addictions, co-morbidities); 2) the context of the first referral to the IOPC program, regarding the course of disease and project of care: dates of first discussion at the OPM, first referral to the PCT and death, oncologic prognosis factors measured within 7 days of first referral to the IOPC program (ECOG [29] performance status (ECOG-PS), serum albumin level, serum C-reactive protein (CRP) level, serum lymphocyte count and serum Lactate Dehydrogenase (LDH) level), elements relative to the course of the disease (at diagnosis, tumour stability or positive response to last treatment, tumour progression), to the project of care (oncologic treatment to come, on course, definitely discontinued or not considered as future option) and to the setting of care (inpatient care, outpatient/ambulatory care); 3) indicators of end-of-life care for decedents: the number of new lines of antitumoral treatment received and length of survival after the first referral to the IOPC program, the place of death, whether the patient had been admitted to a palliative care unit 3 days or less before death, and whether the patient had received antitumoral treatment 14 days before death.

Part of collected data required physician’s expertise (i.e. diagnosis of incurability status) and a good knowledge of patient’s records to be found, most of variables being objective results (i.e. ECOG-PS or lab tests). We therefore had data collected by MDs experienced in both PCT and oncology teams practice (VM), monitored by senior MD researcher and statistician (IC), without actual blind double coding process.

Data were analyzed by description of frequencies (percentage), means (±standard deviations) or medians (interquartile range) as relevant according to the normality of variable distribution and excluding patients with missing data.

In order to investigate the timing of the first referral to the IOPC program, taking into account the pace of progress of the disease, we defined the Index of Precocity (IP), computed for decedents only, as the ratio of the length of survival after first referral to the IOPC program by the length of survival after diagnosis of incurability. Its values lie therefore between 0 (referral to the IOPC program occurs late, close before death) and 1 (referral to the IOPC program occurs early after the diagnosis of incurability). As an example, the IP for a patient with a cancer diagnosed on January 2011 at a metastatic state, referred to the IOPC on March 2011 and deceased on June 2012 is: 15 months / 17 months = 0.88. For a patient with a cancer in metastatic evolution in September 2011, referred to the IOPC on September 2012 and deceased on December 2012, the IP is: 3 months / 15 months = 0.2.

From January 1st, 2011 to December 31, 2013, 445 patients were reported for the first time at OPM. Among them, 416 patients were included for analysis (Fig. 2). Patient’s characteristics at the time of the first referral to the IOPC program are summarized in Table 1.

Patients were mostly men (n = 249/416; 59.9%) with a mean age of 62 years. They presented with some kind of complexity since 63.7% (n = 265/416) of patients had at least one factor of social vulnerability, 28.1% (n = 115/410) had an active addiction and 64.2% (n = 265/413) had at least one serious medical comorbidity.

The characteristics relative to the context of patients’ care are reported in Table 2. At the time of the referral to the IOPC program, 65.2% (n = 270/414) of patients had a progressive disease, whereas 26.6% (n = 110/414) were at diagnosis and 8.2% (n = 34/414) were stable or responded to their last oncologic treatment. Seventy-six percent of patients (n = 314/413) were receiving antitumoral treatment or were about to. The majority of patients received ambulatory care (n = 256/406; 63.1%). When they could be found in health records, collected prognosis factors showed that 44.3% of patients had a ECOG-PS ≥ 3 (n = 176/397), serum albumin levels were < 35 g/l for 54.2% of patients (194/358). A large majority of patients had a CRP > 5 g/l (87.3%; n = 308/353). The overall median [1st - 3rd quartile] index of precocity was 0.39 [0.16–0.72]. The medians of each disease-specific index of precocity are represented in Fig. 3. They range from 0.53 [0.20–0.79] (the earliest, for lung cancer) to 0.16 [0.07–0.56] (the latest, for prostate cancer). The median index of precocity was of 0.51 [0.20–0.76] for bone or soft tissue sarcoma, and of 0.40 [0.14–0.72] for urothelial and bladder cancers.

Whereas the disease-specific index of precocity represents the timing of referral to IOPC, relatively to the duration of the progression of metastatic disease, the Fig. 4 represents a more absolute view of this timing: it represents the median time from incurability to 1st referral to the IOPC program stacked with the median length of survival after 1st referral to the IOPC program. These two representations together provide an overall picture of the timing of referral to the IOPC program, by cancer site.

Indicators of end-of-life care are reported in Table 3 for the 367 patients who died by the end of follow up. Among them, 44.4% (n = 162/365) received one or more new lines of antitumoral treatment after the first referral to the IOPC program. 12.7% (n = 42/332) received antitumoral treatment during their last 14 days of life. Half of the patients died at home or in a palliative care unit (50.7%; n = 186), whereas 37.1% (n = 136) of patients died in an acute care service and 4.6% (n = 17) died in the emergency department or in intensive care units.