Walczak 2017.

Study characteristics
Methods	Aim: to evaluate effects of nurse‐facilitated communication support programme for patients with advanced, incurable cancer to assist them in discussing prognosis and EoL care Study design: RCT, 2 arms; intervention, usual care Consumer involvement: none stated Funding source: NHMRC grant 571346. Authors declaration: no conflicts of interest
Participants	Participants: oncology patients with prognosis of less than 12 months. Advanced, incurable cancer diagnoses of various types, with oncologist‐assessed life expectancy of 2 to 12 months. Informal carers (adult) also participated if nominated by the patient Setting and geographic location: 6 hospital‐affiliated cancer treatment centres; Sydney, Australia Methods of recruitment: oncologists identified consecutive eligible patients at consultations, obtained consent for researcher contact. Oncologists referred patients they expected to die within 12 months but judged likely to live at least 2 months Selection criteria for participation in study: Inclusion: oncology patients with 2 to 12 months left to live (as assessed by oncologist); any cancer type carers: primary, informal providers of care to patient participant in trial. Patients can participate with out a carer, but not viceversa aged over 18 years, capable of consenting to the trial. Both patients and carers must be able to read and write English well enough to complete questionnaires and interviews without an interpreter Exclusion: non‐English speaking cognitive impairment or significant psychological morbidity Diagnosis of person approaching EoL: oncology patients with prognosis of less than 12 months Target of intervention: patients, with or without carer Age: mean 64.4 years (intervention 63.8; control 65.6) Gender: 34.5% female Ethnicity/culture/language: not reported Other PROGRESS aspects: non‐English speaking excluded; psychologically/cognitively impaired excluded Differences between groups on education (higher levels amongst intervention group); more men than women overall in sample; differences in treatments received (chemotherapy rates higher in intervention group); otherwise groups comparable on demographic details. Not clear about representation for other factors Numbers of participants: see Additional Table 1
Interventions	Intervention: communication support programme (CSP) Aim of intervention: overall, to increase patients’ ability and motivation to discuss prognosis and EoL care early in their final year of life (i.e. to assist patients/carers in finding information related to EoL, prognosis, future care, ACP). Both patients’ autonomous motivation and competence were targets of the intervention (i.e. increases in both); and oncologists cued to endorse QPL us and support question asking to address social support needs (‘relatedness’) Comparison:standard care: no contact with nurse, no QPL, oncologists not cued to endorse QPL use or question asking prior to consultations Delivered by: trained senior nurses (experienced in oncology care) 2 nurses; each receiving 40 hours’ training Setting: cancer treatment centres, private environments (education or consultation rooms) Materials, procedures, content: Face‐to‐face nurse meeting; 45 minutes Question Prompt Lists (QPL) introduced by nurse (designed for patients and caregivers with incurable cancer); systematically explored to identify relevant questions Questions included those about prognosis, treatment options and decisions, palliative care, lifestyle, patient and family support, ACP and carer‐specific issues. Prognosis and EoL care issues were highlighted and skills for asking questions discussed Participants also given a DVD on ACP and documenting wishes for care relevant to NSW Participants prompted to identify 1 to 3 questions to ask at next consultation Follow‐up (booster) phone call; 15 minutes 1 to 2 weeks after consultation occurring following the CSP meeting Sought to reinforce content of face‐to‐face meeting and help prepare patients for future consultations using the QPL. Nurses verbally cued oncologists immediately prior to the consultation following the CSP session. Oncologists also received a postcard with suggested endorsement phrasing When and how much: single face‐to‐face session, approximately 1 week before follow‐up oncology consultation. Carers attended where possible Follow‐up telephone booster session 1 to 2 weeks after the consultation following the CSP delivery session Tailoring: tailored as QPL was explored with patients to identify priority questions and to discuss skills for question asking Modified during study: no Co‐intervention(s): none reported Fidelity assessed: fidelity was assessed and shown to be high: assessed after each face‐to‐face and telephone session Key goals completed for almost all participants (see page 34) for CSP Key goals for booster sessions completed for all participants Theoretical base: informed by self‐determination theory of health‐related behaviour change Evidence for effects of QPL and nurse communication support each described; rationale for combining the 2 seems sound (described in more detail in Walczak 2014 study report)
Outcomes	Primary outcomes Information preferences (preferences for amount and type of information) [Knowledge and understanding] Method: Cassileth Information Styles Questionnaire (CISQ); self‐reported questionnaire Timing: baseline; 1 month Scale and scoring: validated and reference provided. Scores subtracted from baseline preference scores; differences expressed dichotomously (preferences met or exceeded if score > 0; unmet is difference < 0) Questions, cues (numbers; from patients, carers) [Discussions about EoL/EoL care] Method: coding of audio‐recorded consultation post CSP Timing: approximately 1 week post‐CSP session (consultation) Scale and scoring: coding scheme developed by authors to identify overall numbers of direct questions and cues for discussion, as well as those relating to specific aspects of care (prognosis, EoL care, future care options, and general issues (latter not targeted by the CSP intervention)) Control preferences (amount of doctor/patient +/‐ carer involvement in decisions) [Evaluation of the communication] Method: Degner Control Preferences Scale (CPS); self‐reported questionnaire Timing: baseline; 1 month Scale and scoring: validated and reference provided. Scores subtracted from baseline preference scores; differences expressed dichotomously (preferences met or exceeded if score > 0; unmet is difference < 0) Patient communication self‐efficacy (PEPPI: Perceived Efficacy in Patient/ Physician Interactions Scale) [Evaluation of the communication] Method: self‐reported questionnaire Timing: baseline; 1 month Scale and scoring: not stated but validated and reference provided Primary outcomes ‐ adverse events None reported Secondary outcomes Health‐related quality of life [Quality of life] Method: health‐related QoL (FACT‐G). Self‐reported questionnaire Timing: baseline; 1 month Scale and scoring: validated and reference provided; other details not stated Consultation length [Health system impacts]
Notes	Satisfaction with the face‐to‐face session and with follow‐up call were also reported for the intervention group but as data were not comparative it was not included in the review Numbers of questions and cues were also reported; for analysis only numbers of questions were used
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Computer‐generated random number table was used to generate blocks of 1:1 balanced randomization codes for each referring oncologist" (page 32)
Allocation concealment (selection bias)	Low risk	Allocation sequence concealed in sequentially numbered, opaque envelopes; sequentially opened by blinded research manager for oncologist to determine randomisation
Blinding of participants and personnel (performance bias) All outcomes	High risk	"Participants and oncologists could not be blinded" (page 32) Questionnaire measures are self‐reported by patients/carers: lack of blinding may affect these ratings, particularly for several of the subjective outcome ratings
Blinding of outcome assessment (detection bias) All outcomes	High risk	"Participants and oncologists could not be blinded" (page 32) Questionnaire measures are self‐reported by patients/carers: lack of blinding may affect these ratings, particularly for several of the subjective outcome ratings No information on whether those coding consultation recordings for analysis were blinded to group allocation
Incomplete outcome data (attrition bias) All outcomes	High risk	Attrition was high (31/110 (28%) lost to follow‐up), possibly largely explained by declining health of participants (patients) Higher in intervention group (34% intervention group, 18% control). No systematic reasons for differential attrition were identified by authors but dropout in intervention group is substantial and may introduce bias Authors state that ITT analysis was used (according to group assignment) but dropout rates may be problematic
Selective reporting (reporting bias)	Unclear risk	Checking against the protocol in Walczak 2014 there are a number of outcomes not reported (e.g. acceptance of disease, preferences for future interventions, etc,) but primary outcomes are reported. It is possible that other outcomes are reported elsewhere
Other bias	Low risk	Groups were similar at baseline (other than higher educational levels and rates of chemotherapy in intervention group) Differences between groups on education (higher levels amongst intervention group); more men than women overall in sample; differences in treatments received (chemotherapy rates higher in intervention group); otherwise groups comparable on demographic details. Not clear about representation for other factors Contamination is possible as patients were the unit of randomisation (rather than at the level of the oncologist). Not clear if this is likely however Trial is underpowered (sample size calculated at 140; 110 recruited; 79 completed) to detect differences between groups

ACP: advance care planning; COPD: chronic obstructive pulmonary disease; CPR: cardiopulmonary resuscitation; EMR: electronic medical record; EoL: end of life; ICC: intracluster correlation; ICU: intensive care unit; IQR: interquartile range; ITT: intention to treat; QALY: quality‐adjusted life‐year; QoL: quality of life; RCT: randomised controlled trial; STAI: State‐Trait Anxiety Inventory; SD: standard deviation; UC: usual care.