Introduction
Design
Systematic review
Search strategy | Electronic searches of four research literature databases (PubMed, EMBASE, Cochrane (SR & Trials), Web of Science (SCI)) were conducted from 2000 up to 15 July 2015. In addition, searches for grey literature were conducted in OAIster, OpenGrey, NYAM Grey Literature Report and Lexis Nexis, up to 11 September 2015. Details of the electronic searches are provided in supplementary resource S2. Searches of eligible studies’ reference lists and forward citation tracking (using Google Scholar and PubMed) were also conducted. Initially, no language or date restrictions were applied, but due the large number of references identified, the search was restricted to English language publications and to full publications (i.e. not conference abstracts) |
Study selection and data extraction | Retrieved title–abstract records were initially screened by one of several reviewers (SK, JE, SB, TB-G or EH) against the PICOS criteria. A random sample of 10 % of records was independently screened by a second reviewer. Full-text screening of potentially eligible study reports was undertaken by at least two reviewers working independently, with any discrepancies resolved by a third reviewer. Study data were extracted by one reviewer and checked by a second reviewer |
Risk of bias assessment | Risk of bias assessments were done by one reviewer and checked by a second reviewer. For the RCTs, the Cochrane Risk of Bias tool [15] was used to assess potential bias in studies for each outcome; six criteria were considered: random sequence generation, allocation concealment, blinding of participants, personnel (where feasible) and outcome assessors, baseline comparability between groups, incomplete outcome data, and selective outcome reporting. Risk of bias judgements in domains judged most critical (sequence generation, allocation concealment, incomplete outcome data) was used to determine a summary study-level risk of bias. For cohort studies, and pre–post studies, we used criteria published on websites [17, 18]. Lastly, we used GRADE criteria [19] to enable assessment of the quality of the body of evidence for each outcome |
Key informant interviews
Stakeholder consultation at the IBTA summit
Results
The impact of QoL assessment tools in routine practice in oncology settings
Physician (or nurse)–patient communication
Patient management
Outcome | Overview of results |
---|---|
Number of actions taken for specific domains/issues (n = 3 studies) | Of two studies that reported on the number of actions taken for specific domains, only Nicklasson et al. [23] reported that the number of diagnostic and therapeutic interventions directed to emotional concerns (SMD 0.45 [95 % CI 0.15–0.76]) and social concerns (SMD 0.38 [95 % CI 0.08–0.69]) were larger in the intervention group compared to the control group. In addition, a study by Wolfe et al. [24] reported that the QoL intervention led to the initiation of a psychosocial (56 %), pain (34 %), social work (33 %) or palliative care (29 %), consult, but more detailed data were not reported |
Number of specific actions (n = 4 studies) | No studies reported differences between intervention and control groups for prescription of medications [21, 30], ordering of tests [21, 30], referrals [21, 35] or modification/cessation of chemotherapy [21]. Of three studies that evaluated advice/counselling as an outcome [21, 22, 30], only Detmar et al. [30] found that a statistically greater percentage of patients in the intervention group received counselling from their physician on how to manage their health problems compared to those in the control group (23 vs. 16 %) |
Category | QoL tool name |
---|---|
Brain cancer specific | EORTC QLQ-BN 20 |
FACT-br | |
FACT-mng | |
Patient concerns inventory | |
M.D. Anderson Symptom Inventory-Brain Tumor Module | |
Cancer specific | Distress thermometer |
Macmillan holistic needs assessment | |
Distress Assessment & Response Tool (DART) | |
Psychological Screen for Cancer (PSCAN) | |
General | EQ-5D |
SF-36 | |
Karnofsky Performance Status scale | |
Barthel Index |
Facilitators | Challenges | |
---|---|---|
System-level factors | A formal requirement for use in trials can influence care of patients participating in trials and patients in settings where trials take place Availability of human resources (e.g. trained nurses and doctors to administer and interpret findings) Value placed on patient perspectives and patient engagement by the health system Although rare, the presence of guidelines on tool usea
| A general lack of policy and guidelines for the use of QoL instruments in cancer care contexts A need for training of health professionals, confounded by health system resource constraints and time demands Challenges to effectively communicating QoL findings from clinical trials to the point of prescription A lack of insights on the place of QoL assessment findings in valuations of drugs |
Tool-related factors | A relatively simple design is important in a care context (with simplicity being relatively less important in trial contexts where the capacity and resources to cope with more detailed instruments is better established) Ability of questions to address aspects of QoL that are meaningful to the patient and the clinician (this are not mutually exclusive but not always the same in priority) Scope for remote administration (e.g. iPads, telephone administration) but requires reliable devices, software and training | No “gold standard”: perceptions of overly complex or overly simple designs and the associated need for some standardisation in trials but sufficient tailoring for care contexts (across patient profiles and cancer types and stages) Cultural specificities associated with HRQoL and implications on the nature of questions that need to be asked Having patients complete (and staff process) different forms for different cancers was seen to potentially be too burdensome |
Administration, data and disease-related factors | Follow-up on findings with patient/carer as key for public acceptability of tools and their take-up | A general lack of awareness about the diversity of available tools and how to access and engage with them, among cancer patients |
Scope for proxy-reporting (but not without trade-offs) when patients are unable to complete questionnaires directly (e.g. due to issues such as cognitive decline) | Shifting patient expectations of HR QoL during the course of disease Patient versus proxy views can be different |
Current use of QoL assessment tools in healthcare delivery for cancer patients, with particular emphasis on brain cancer
Tool used in care delivery, as reported in interviews | Use |
---|---|
Distress Thermometer | Used in the care of patients with a variety of cancers including brain cancer, in some neuro-oncology settings in the Netherlands. Dutch guidelines for psychosocial care of cancer patients recommend the use of QoL measures, and, as one interviewee noted, although it “… is not a real QoL instrument… it makes it possible to discuss QoL with patients… and can help to detect complaints and issues”. (INT04). This in turn was seen to facilitate more appropriate referral pathways |
Distress Assessment and Response Tool (DART) | Mandated by the provincial government of Ontario as a general cancer QoL measure for assessing all patients in cancer centres in Ontario, Canada. It is administered each time a patient visits the clinic (unless they are undergoing daily radiation treatment). The data are used by clinicians who scan the results before meeting patients and can ask patients specific questions that warrant attention in follow-up |
Psychological Screen for Cancer (PSCAN) | Used by the British Columbia Cancer Agency in Canada, as part of its Patient and Family Counselling Services Initiative, as it covers issues they feel counselling can help with. It is currently administered to all patients at their first visit, and they are trialling using it repeatedly during a patient’s care |
Barthel Index | Used in clinical practice in some settings in Spain to gather information on the status of a brain cancer patient, in terms of performance in activities of daily living |
Interviewees also suggested that QoL tools could help empower patients to play a more informed role in decision-making through a greater understanding of the implications of particular treatment on their QoL. This was considered to be particularly pertinent to brain cancer care, given that treatment is not curative.14, 15“QoL is an important outcome as it tells you how the patient feels and we know there is a difference between clinical parameters and subjective patient parameters. Someone can be terribly sick but may not feel it that way and vice versa. QoL is about the patient perspective and the functioning of the patient in his or her entirety”.13
“There is a particular concern with QoL in the brain tumour field because there are no cures for many of the diseases covered. Where there is no cure, the priority is to minimise harm done to the patient by treatments”.16
Factors influencing the use of QoL assessment tools in cancer care, with a particular emphasis on brain cancer
The subjective nature of QoL assessment also affects ease of interpretation, as different patients will differ in their expectations of QoL and expectations may also alter as the disease progresses, so tools may not remain equally applicable across disease stages.29“Training is particularly important, as it is difficult to interpret patients’ responses – clinicians are used to looking at physical symptoms and find it challenging to adapt to incorporating patients’ assessment of their symptoms into their decision-making”.28