Quality of life versus length of life considerations in cancer patients: A systematic literature review

1A diagnosis of cancer can be devastating, and deciding on the appropriate treatment can be complicated and daunting. Patients are asked to consider factors that include mortality from the disease and the potential for acute and chronic morbidity from the treatment. Appropriate decision making requires satisfactory patient understanding of these treatment choices, which includes the potential benefits and harms.1 The primary focus of cancer treatment has always been to increase overall and disease free survival; however, quality of life (QoL) has been increasingly recognized as an important end point.2

2Although there is an instinctive understanding of the term “quality of life,” there are multiple definitions, which gives testimony to the fact that it is a complex concept with many diverse facets and components. The standard dimensions used in QoL questionnaires measure the presence or absence of specific symptoms or overall general health. They do not measure patients' beliefs or attitudes toward treatment and intervention outcomes.3 Decision making in a cancer setting can be a difficult process due to its multifaceted nature. The patients' outlook and beliefs are paramount, but this is heavily influenced by their own experiences and those of friends and family.4 In addition, current QoL and physical status can affect subsequent decisions.

3Most cancer trials primarily focus on the standard oncology end points relating to survival, but it is possible to derive composite measures, which assess the impact of QoL on the final outcome of different therapies. These are called quality adjusted survival metrics or health utility metrics, and a wide range of them have been developed over the past 30 years. Utility measures allow patients a chance to value a different perspective on treatment and outcomes. Two methods of utility measurement that may be used to calculate quality adjusted life years (QALY) or quality adjusted survival are standard gamble and time trade‐off (TTO).5 In standard gamble, patients are asked to choose between staying in a state of ill health for a specified time period or choosing a treatment that may either cause their death or restore perfect health. In the case of TTO, the individual expresses a preference between two choices, usually between LoL or a better health status.4 These methods have been increasingly adapted in cost‐utility analyses of pharmaceuticals and various health‐care interventions. In reality, scenarios are often more complex with disease and treatment effects impacting variably on QoL over a prolonged time course. There may be a significant drop in QoL after an intervention but an overall better long‐term QoL and increased life expectancy. QoL measurement should not just focus on a single time point when assessing an intervention.

4In cancer treatment, patients are often required to make trade‐offs between QoL and length of life (LoL).6 Tumor‐specific therapy can potentially prolong life; however, this may reduce QoL significantly. Some patients are willing to endure toxicities associated with treatment in order to increase their LoL, while others value QoL more and are reluctant to spend their remaining years in a compromised state.7 This involves weighing the risks and benefits of treatment and managing the patients' concerns and expectations. There may be personal reasons associated with their health, the effect on their family and friends, and the consequences of the treatment itself. A trade‐off for potential gain in life expectancy may involve short‐term debility from treatment (postsurgical pain, chemotherapy‐induced nausea and alopecia, and etc) or permanent side effects (stoma, disfigurement, physical dependency, and etc). Moreover, the compromise is not always related to health but instead may be about financial burdens and increased dependency on friends and family.

5To understand cancer treatment choices concerning trade‐off, various questionnaires and methodologies have been devised to understand patient preferences and priorities toward cancer treatment. Quality‐adjusted time without symptoms or toxicity (Q‐Twist) allows the combination of both quality and quantity of survival time.8, 9 The principle hypothesis of this method is that patients without disease symptoms or treatment toxicity have a better health‐related quality of life (HrQoL) than those who have disease‐specific symptoms and toxicity. Q‐TWiST was initially used to assess adjuvant therapy for breast cancer and has now been adapted in other cancers.10, 11, 12 The Quality/Quantity Questionnaire designed by Stiggelbout and colleagues was created to assess patients' preferences toward either QoL or LoL when deciding about cancer treatments.7 Other methods include discrete choice experiments and various bespoke questionnaires tailored to a specific study.13, 14, 15

6The aim of this review was to determine the factors influencing patient preferences for either QoL or LoL and how these impacts on cancer treatment choices.

7A systematic literature search was performed according to PRISMA guidelines (see supporting information) using five databases between 1942 and October 2018. The databases included MEDLINE, SCOPUS, Cumulative Index to Nursing and Allied Health Literature (CINAHL), PsychINFO, and Web of Science. A pilot search on MEDLINE, was performed to identify the relevant keywords contained in the title, abstract, and subject descriptors. Five broad categories of concepts were searched: “quality of life,” “cancer,” “length of life,” “health utilities,” and “decision making.” The search terms included (cancer* OR neoplasm* OR oncolog* or tumo?r*) AND (quality of life OR QoL) AND (Longevity OR Length of Life) AND (decision making OR patient participation OR patient preference OR patient participation OR treatment choice) AND (health state utilit* OR standard gambl* OR trade‐off). See Appendix S1 for the search strategy as used in Ovid Medline. The literature search was carried out by two authors (A.S. and C.M.).

8A study was only included if there was reference made to preference for QoL or LoL with or without determinants that may influence treatment choice. These factors could be either demographic influences, health status, or personal factors. Study designs could be qualitative, quantitative, or of mixed methods. Studies included were limited to adults with cancer and published in English. A PRISMA format was used to filter through articles. Editorials, reviews, and expert opinions were excluded. Hypothetical studies with healthy volunteers were also excluded as it was felt that these studies were unrealistic in their assessment of whether LoL or QoL would be favored in a cancer setting. Health status utilities were included in the search to include any trade‐off papers suitable for review. Time trade‐off studies may indicate treatment preferences, however not necessarily in the context of a preference for QoL versus LoL. Only those focusing on QoL versus LoL preferences were included.

9Study selection was by a two‐step process by two independent reviewers (A.S. and C.M.), at titles and abstract stage with arbitration for articles with uncertainty. In the second stage, full‐text articles were independently reviewed (Figure 1). Reference lists of all selected articles were reviewed to identify any additional relevant articles, identifying five further articles. When an article referred to additional publications for more details concerning study methods and design, those publications were also acquired.

10Data extraction was performed by two independent reviewers (A.S. and C.M.). The information collected included study design, aim of study, location of study, sample size and response rate, age of the sample, type of cancer, any research tools used in the form of questionnaires and the findings of the study relating to QoL versus LoL preferences.

11The Mixed Methods Appraisal Tool (MMAT) was used to quality assess the articles that were included in the study. The 2011 MMAT tool encompasses five types of mixed methods study components or primary studies: qualitative, quantitative randomized controlled trials, quantitative nonrandomized, quantitative descriptive, and mixed methods, each with its own set of methodological quality criteria. For each item the response categories were “yes,” “no,” or “can't tell” followed by comments.16 Higher quality is denoted by the number of stars (*) in the tables. Quality assessment was independently scored by two reviewers (A.S. and C.M.). No study was excluded based on quality assessment, as all were of acceptable quality.

12The literature search revealed 4388 articles. A total of 843 abstracts were excluded because of duplication, and 3494 articles were declined as they were either reviews, expert opinions/editorials, or not suitable for the topic under review. A total of 56 articles were reviewed fully, and only 30 deemed suitable for inclusion. The 26 rejected papers were not suitable as they were either reviews or not relevant (Figure 1). Included studies are summarized in Tables 1 (quantitative), 2 (mixed methods), and 3 (purely qualitative) (Tables 2 and 3).

13The majority of studies identified in this review were quantitative. Generic questionnaires (EORTC‐QLQ‐C30 and FACT‐G) and disease specific questionnaires (EORTC‐QLQ‐H&N) were used to assess QoL. The studies were mainly conducted to understand the decision‐making process in the advanced cancer setting. The studies had wide focus that included understanding the role of the doctor and the attitude the patient has toward their treatment, among other themes. Understanding QoL and LoL trade‐offs as part of the decision‐making process, usually formed a limited part of many of these studies.

14Meropol and colleagues (2008) suggested that QoL and LoL are both equally important; however, the majority of patients with advanced cancer in this study prioritized QoL over LoL.41 This was also reflected by the study of Jenkins and associates.36 Silvestri and associates noted although there were some patients who would endure treatment and associated toxicities just to live a single day longer, there were also patients who would decline all treatments. These latter patients would rather maintain their QoL and having to withstand the adverse effects of treatment would not be a worthwhile trade‐off.20 The authors postulated that patients may opt for enhanced QoL only if the chance of survival was less than 50% relative to baseline survival (without treatment).42

15Many patients in the study by Brom and colleagues felt that they ought to have some sort of intervention for their cancer and found it difficult to accept the concept of LoL and QoL. Although some patients opted for treatment initially, they expressed the view that if it was affecting their QoL, they would cease treatment.39 Marta and colleagues noted that the majority of patients in their study wanted to undergo a treatment that would prolong life but not compromise their QoL.43 In a qualitative study by Gerber and colleagues, patients stated that they were keen to maintain their activities and not be a burden on family, and therefore not undergo chemotherapy if those factors were compromised, indicating the importance of QoL.38

16Survival seemed to be a key feature in the decision‐making process and patients were found to opt for treatment if they felt that their prognosis was likely to improve.15, 19, 28, 40 Their current health status also affected their choice. Perez and associates found that those who wanted to trade time, scored lower in many of the domains of the baseline HRQoL questionnaires.3 Patients in better health were found to rate LoL more highly, whereas those who were in poorer health strived to maintain their QoL.7, 22, 32, 44 Kiebert and associates noted that issues patients felt were important were baseline QoL and the probability of survival.17

17Kiebert and associates assessed factors affecting decision making for cancer treatment and noted that important factors were age, marital status, children, inability to work due to side effects, disease related life expectancy, and baseline QoL. No significant associations were found between the various determinants; however, patients did rate having children and marital status as somewhat important in decision making.17

18Other studies have shown different results, with gender, children, education, religion, and cancer type not influencing treatment choices.3, 6, 23, 35 Those with strong family links preferred survival. Unemployed patients prioritized QoL.6 Wong and colleagues concluded that those who were able to pay for their treatment chose to have treatment to prolong their life.45 These latter findings are only relevant in self paying health care systems.

19Many of the studies carried out have not been age specific; therefore, it has been difficult to make inferences about the influence of age on LoL/QoL preferences. The studies in this review show a mixed picture. Older patients have a preference for QoL, which is not surprising considering natural limitations to life expectancy and the often reduced QoL associated with advanced age.34 Younger cancer patients were more likely to tolerate aggressive treatments to increase survival years.30, 35, 46 A study by Pisu and colleagues involving 170 ovarian cancer patients, showed that maintaining QoL and living as long as possible were both important. In women less than 65 years old, 96.9% felt longevity was important, and 95.9% felt that preserving QoL was important, compared with 87.5% and 90.3%, respectively, in the greater than 65‐year‐old age group.33 Stiggelbout and associates noted that when age was adjusted for in their statistical calculations, those in relationships and with children preferred longevity.7 Derks and colleagues found that older patients were less likely to receive standard treatment, an effect that was more evident in those above the age of 80 years old. Reasons behind this included lack of social support and being widowed. Patients who did not receive standard treatment also prioritized QoL more strongly.27

20When looking at symptom tradeoffs against longevity, patients were prepared to tolerate certain treatment side effects to live longer. Patients were willing to prioritize survival over intact sexual function in prostate cancer for instance.18, 44 When patients with advanced cancer reached the end of their lives and had to endure pain and discomfort, 47% of patients chose to have palliative surgery to maintain or enhance their current health status and independence.37

21Patients suffering from cancers with a good prognosis such as breast and testicular cancers, compared with recurrent colorectal or lung cancer had similar thoughts regarding QoL and LoL.7 Despite the type of cancer, patients felt that QoL and LoL were equally important when considering treatment.41 In the study by Pisu and colleagues involving ovarian cancer, more than 90% stated that QoL and LoL were equally important.33 Another study by Jenkins and associates, involving participants with ovarian cancer showed that 57% felt LoL and QoL were equally important, 9% prioritized LoL, and 33% favored QoL.36 However, Donovan and colleagues demonstrated that women who had recurrent ovarian cancer, would opt for LoL, and choose to receive aggressive treatment, QoL was a secondary issue.23 Patients with a shorter history of cancer preferred LoL; however, those with poorer prognosis and closer to their predicted time of death valued QoL more.35 In contrast, Meropol and colleagues found that there was no association between time since diagnosis and QoL/LoL preference.41

22This study presents the first comprehensive review of studies looking at trade‐offs between QoL and LoL in a cancer setting. The aim of this review was to highlight whether patients prioritize QoL or Lol and the determining factors that influence the decision‐making process for cancer treatment. In fact, the findings indicate that many of the studies do not directly test determinants. The QQ questionnaire has been designed specifically to quantify the patient's choice of QoL or LoL and also to what extent patients would be inclined toward either. The questionnaire does not capture the psychological reasoning behind the preference however. It is also perhaps more suited for patients with advanced cancers where the cancer will inevitably cause death regardless of whether it was treated or not.7 For some patients, where curative treatments may be available, albeit with a high cost (for example, mutilating operations leading to disfigurement, ie, head and neck resections, mastectomy, and amputations) or where death due to old age or other, noncancer comorbidities is imminent; this trade‐off may also be relevant and the QQ tool is not designed to explore these scenarios.

23This review highlights the importance of carrying out baseline QoL assessments prior to treatment and evaluating the impact of life expectancy. The importance of performing age specific studies is also noted as priorities between younger and older patients are different. The preferences for QoL or LoL by younger patients, may be influenced by their desire to spend time with their partner or children. Older patients are more likely to suffer from multiple comorbidities and be frailer, and discussions may need to include whether a treatment will be tolerated less well because of these limitations, or result in an increased risk of harm. Considerations should include patient intolerance to certain chemotherapy agents or surgery, as well as an understanding that they may never reach their preoperative baseline physical fitness again after treatment. This “step down” in function tends to be more prominent in the older age group,47, 48 an effect that is widely recognized across many medical interventions in older patients. They may feel that time spent receiving treatment may not be worth the extension of life for a relatively short period. Older individuals have a good overall understanding that they have lived their lives and are more accepting of the inevitability of death and of their physical limitations. Studies suggest that a good QoL in older people is often based around the following: independence, a strong social circle, and an ability to retain their “inner selves.”49 These values may be compromised by having treatment. Other studies have shown that the most consistent factor influencing treatment decision making in older patients is a recommendation from doctors.50 In breast cancer, undertreatment is well‐documented in older patients.51 This has led to avoidable disease‐specific deaths.52 Exploring the patients' views regarding treatment at an early stage would help reduce the impact of age‐related clinician bias, which is well recognized.53

24Decision making in cancer treatment is difficult as there are multiple components to consider aside from the purely medical aspects. Likewise, the compromises the patient is willing to make can vary greatly depending on many factors including patient age, personal family dynamics, social structures, and, patients' likely survival and baseline QoL. This may subsequently impact on whether the patient is more inclined towards longevity or QoL. Although there are studies trying to understand the factors influencing the final decision, there is limited information on preferences between QoL and LoL and the trade‐off the patient is willing to make. Clinicians have influence over the final decision, and therefore it is vital for the patient to have a full understanding of their treatment and the impact it may have on their life.

25This study is the first to use a rigorous and systematic approach to review studies based on patient preferences regarding QoL or LoL in a cancer treatment setting. Despite a comprehensive database search strategy, it is possible that some relevant articles may have been missed and despite the various methodologies, all papers included were of an acceptable design and standard for inclusion. However, the main findings of the review are likely to be robust to missing studies. On the basis of our interpretation and weighting of the evidence, we are confident in the conclusions that have been drawn from findings across several studies rather than be based on isolated studies. None of the studies in this review has looked at the impact of preexisting, noncancer‐related limitations to life expectancy as part of this trade‐off, such as is seen in the oldest age groups and the impact of acceptance of impending age‐related mortality. With the aging of Western populations, this is an important gap in the literature.

26The studies included in this review are exploratory cohort studies carried out in a retrospective manner, whereby patients have already made their decision regarding treatment. There may be a source of bias influencing their responses, as many issues may not have been considered prior to treatment or the decision‐making process.

27Many of these studies have mainly focused on advanced cancers of all types. For patients who are facing mortality imminently, the decision to prioritize QoL and LoL is pertinent. In the case of slow growing cancers such as prostate and breast cancers, where conservative management is widely accepted, the choice between QoL and LoL can be more complicated. Patients often die from other causes rather than the cancer itself.54 As the majority of the articles identified in this search did not involve early stage cancer, it is difficult to know what patients envisage from their treatment and what trade‐offs they were willing to make as well as how these factors may change with the course of the natural disease process. This is where patients' age and comorbidities may play a larger role in whether the patient opts for QoL or LoL.

28This review has several important clinical and research implications. With treatment and care now becoming more patient centered, it has become more pertinent to understand the impact of the cancer diagnosis on the patient and the motivations behind their treatment choices. The impact of treatment of certain cancers may be extreme and may involve a great deal of compromise and acceptance of change in circumstances. Factoring the likely impact of treatments on QoL relative to that at baseline should be discussed with every patient. This would ensure that patients have a full understanding of what their treatment entails and that they are aware of the consequences of treatment and nontreatment. Further in‐depth studies are required to understand the emotional and physical considerations and personal priorities the patients may have during the decision‐making process. This may go a long way in elucidating what aspects of their life they are willing to trade to maintain their QoL or increase LoL. Older age specific issues and cancer specific decision‐making processes also need exploring.

29The authors have declared no conflicts of interest. The views expressed are those of the authors and not those of the NHS, the NIHR, or the Department of Health.