If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding authors at: Cicely Saunders Institute of Palliative Care, Policy and Rehabilitation, King’s College London, Denmark Hill, London, SE5 9PJ, UK. (M. Maddocks). Occupational Therapy Department, Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, 308433, Singapore, (J. Neo).
Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, 308433, SingaporeCicely Saunders Institute of Palliative Care, Policy and Rehabilitation, King’s College London, Denmark Hill, London SE5 9PJ, United Kingdom
Corresponding authors at: Cicely Saunders Institute of Palliative Care, Policy and Rehabilitation, King’s College London, Denmark Hill, London, SE5 9PJ, UK. (M. Maddocks). Occupational Therapy Department, Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, 308433, Singapore, (J. Neo).
Activities of Daily Living (ADLs) are essential to live independently within society.
•
One-third of adults with cancer have difficulty performing basic ADLs.
•
Half of adults with cancer require assistance to perform instrumental ADLs.
•
The most commonly affected ADLs include walking and transfers, housework, shopping and transportation.
Abstract
Introduction
People with cancer frequently report limitation in Activities of Daily Living (ADLs); essential activities required to live independently within society. Although several studies have assessed ADL related disability, variability in assessment, setting, and population means evidence is difficult to interpret. We aimed to determine the prevalence of ADL related disability, overall and by setting, and the most commonly affected ADLs in people living with cancer.
Methods
We searched twelve databases to June 2016 for observational studies assessing ADL disability in adults with cancer. Data on study design, population, ADL instruments and disability (difficulty with or requiring assistance in ≥1 activity) were extracted, summarised, and pooled to estimate disability prevalence with 95% confidence intervals (95% CI) overall and by setting.
Results
Forty-three studies comprising 19,246 patients were included. Overall, 36.7% (95% CI 29.8–44.3, 18 studies) and 54.6% (95% CI 46.5–62.3, 15 studies) of patients respectively reported disability relating to basic and instrumental ADLs. Disability was marginally more prevalent in inpatient compared to outpatient settings. The Katz Index (18 studies) and Lawton IADL Scale (11 studies) were the most commonly used instruments. Across the activities studied, the most frequently affected basic ADLs were personal hygiene, walking and transfers, and instrumental ADLs were housework, shopping and transportation.
Conclusions
About one-third and half of adults with cancer respectively have difficulty or require assistance to perform basic and instrumental ADLs. These findings highlight the need for rehabilitation focused on functional independence, and underscore the importance of professionals skilled in occupational assessment and therapy within cancer services.
With advances in screening and therapies, people are increasingly living longer with the consequences of cancer and its treatment. Patients with cancer frequently report a sustained symptom burden, exercise intolerance, and physical deconditioning, all of which can threaten everyday independence [
]. Activities are categorized into basic ADLs; referring to personal care activities such as feeding, toileting, washing, and dressing, as well as instrumental ADLs; referring to extended tasks such as meal preparation, using public transportation, doing household chores, and grocery shopping [
]. Disability relating to ADLs (difficulty with or requiring assistance in at least one activity) has been associated with poorer quality of life, and in older adults is predictive of mortality [
]. It follows therefore, that where disability relating to ADLs exists, addressing it should be a core goal of clinical management.
Although multiple studies assessing ADL disability in cancer have been published, the variability among study populations, settings, and measures makes it difficult to interpret the evidence. Landmark studies have used ADL disability to characterise functional decline at the end of life, for example Lunney et al. [
] demonstrated more varied trajectories of disability. Whilst these studies highlight the severe impact of cancer on ability to manage ADLs, both were limited to a community setting and late phase of illness. Earlier in the cancer trajectory, individuals tend to experience more disability in instrumental ADLs compared to basic ADLs [
]. Evidence regarding which activities are most commonly affected in cancer questions the original hierarchy of ADL proposed by Katz. For example, both eating [
To our knowledge, no review has systematically examined ADL-related disability in adults living with cancer. A better understanding of disability profiles and patterns regarding ADLs, including in the context of cancer stage and settings, would allow clinicians to target rehabilitation interventions, shape policy around rehabilitation services, and equip researchers with a knowledge base on which to develop treatments to reduce ADL disability. This review aimed to identify instruments, items, and reporting of ADL disability in people with cancer; to determine the prevalence of ADL disability overall and by setting; and describe the most commonly affected basic and instrumental ADLs in this group.
Method
Design
The systematic review was planned in accordance with the Centre for Reviews and Dissemination [
(CRD) Cfrad. Systematic reviews: CRD's guidance for undertaking reviews in health care. York (UK): Centre for Reviews and Dissemination, University of York; 2009.
] guidance on conducting reviews and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [
Studies of adult populations (≥18 years), with a primary clinical or histological diagnosis of cancer were included. Studies of paediatric populations (<18 years) were excluded as they are recognized as a distinct population with different ADL requirements from adults. Cancer survivors were also excluded as the term survivor has been defined in multiple ways, from those living with cancer to those cured and free of any disease [
]. Studies employing any measure of ADL disability were included. We did not consider studies only employing measures of global functional status, e.g. ECOG Performance Status, or physical limitation, e.g. walk tests. All prospective and retrospective observational study designs were eligible, including cohort studies, cross-sectional studies, longitudinal studies, case series, and chart reviews.
Search strategy
An electronic search strategy using a combination of full-text search terms and MeSH terms was developed for MEDLINE and adapted where necessary for all other databases. Search terms included “cancer” or “carcinoma” AND “functional disabilities”, “functional outcomes” or “functional impairment” AND “observational studies”, “cohort studies” or “longitudinal studies” All search terms were used as full-text, with use of truncation symbol to retrieve variations in the terminology. (Supplement, Appendix 1.)
A systematic literature search was conducted in 12 electronic databases from inception to June 2016: MEDLINE; EMBASE; CINAH; ASSIA (Applied Social Sciences Index and Abstracts); PsychINFO; Social Policy and Practice; IBBS (International Bibliography of the Social Science); Science Direct; Social Service Abstract; Sociological Abstract; Scopus; Web of Science Core Collection. Grey literature was searched on six databases: OpenGrey (System for Information on Grey Literature); ProQuest Dissertations & Theses; Web of Science Conference Proceedings; Scopus Conference Proceedings; HMIC (Health Management Information Consortium); Global Health. No time restrictions were applied. Restrictions were applied to studies in human subjects published in English language.
Search alerts were set up regularly for updates of relevant new publications in each database up to November 2016, and hand searching was conducted up to October 2016. The first author (J.N.) scanned reference lists, and performed forward and backward citation tracking of included studies. Experts in the field of cancer and disability were also contacted via email to seek potentially relevant research material, including ongoing and unpublished research.
Data collection and analysis
Selection of studies
An online reference management system (Refworks) was used to manage electronic database hits and remove duplicates. Eligibility criteria were first applied to the title and abstracts. Full-text articles were retrieved for titles/abstracts that meet the review criteria or when information in the title and abstract was insufficient to determine eligibility. The appraisal of articles against inclusion and exclusion criteria was agreed between the first (J.N.) and last (M.M.) authors, with any disagreements regarding the studies resolved via discussion.
Assessment of methodological quality in included studies
Studies selected for inclusion of specific items of ADL disability were assessed for methodological quality using Loney’s critical appraisal tool [
], designed to assess quality and susceptibility to bias in articles determining the incidence or prevalence of a clinical issue. The tool covers eight criteria; description of study subjects, random sample, unbiased sampling frame, adequate sample size, standardised measures, unbiased assessors, adequate response rate, and confidence intervals and subgroup analysis. A point is allocated for the presence of each criterion. A higher score indicates higher methodological quality.
Data extraction and analysis
A standardised data extraction form was developed, piloted, and data from eligible studies was extracted (J.N. and M.M) to retrieve data on study design, setting, sample characteristics, instruments of ADL disability, and characteristics of ADL related disability.
Normally distributed variables were summarised by mean and standard deviation. For studies with multiple subgroups within the sample, data were aggregated into a single group using the formula devised by the Cochrane Collaboration for combining means and standard deviations of groups [
]. When the range was not available for calculation of mean, median was assumed as mean for studies of sufficiently large sample size. Categorical variables e.g. cancer type, were summarised by percentage of participants in the highest occurring category.
Data on overall and specific ADL disability prevalence were summarised by mean and 95% confidence intervals calculated using Wilson’s method (Confidence Interval Analysis version 2.1.2 software). Disability prevalence was determined overall, and by inpatient/outpatient setting as patients admitted to hospital or care home might be more dependant in ADLs. Data on number of ADL disability and total ADL disability score were summarised as mean and standard deviations. Figures were produced using Graphpad Prism software 7.0b. A meta-analysis was conducted to synthesise the results from multiple studies into a single prevalence point for disability relating to basic and instrumental ADLs, overall and by setting, using Cochrane Collaboration’s RevMan software [
Our search of the literature retrieved 13,432 articles. After de-duplication and title/abstract screening, 83 full-text articles were retrieved for further appraisal, of which 43 separate studies were included (Fig. 1, see Supplement, Appendix 2 for excluded studies). All studies were included for narrative review of instruments and items of ADL disability. Data from eighteen and fifteen studies respectively were used for meta-analyses for pooled estimates of basic ADL and instrumental ADL disability.
Characteristics of included studies and sample participants
See Table 1 for information on characteristics of all included studies. Most studies were conducted from 2000 onwards (n = 32), in the US and Canada (n = 21), Europe (n = 19), or Asia (n = 3). The majority of studies (n = 24) recruited participants from outpatient settings, e.g. home care, clinics, or community-dwelling adults. Data from a total of 19,246 participants were available, with individual study samples sizes ranging from 45 to 6822. Participants were heterogeneous with regards to primary cancer type (breast, colorectal, lung, gastrointestinal, etc.) and cancer treatment status.
Table 1Characteristics of included studies for narrative review.
End-of-life care during and after an acute hospitalization in older patients with cancer, end-stage organ failure, or frailty: a sub-analysis of a prospective cohort study.
Comprehensive geriatric assessment adds information to Eastern Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study.
Instruments, items and reporting of ADL disability
In all studies, ADL disability was defined as having difficulty with or requiring assistance in at least one activity. Most studies used standardised and validated instruments, however, close to one-quarter of studies (n = 9/43) used questionnaires and/or scales constructed specifically for the study, with items from a parsimonious selection of ADL. The most commonly used ADL instruments were the Katz Index of Independence to assess basic ADLs (n = 18), and the Lawton IADL scale (n = 11) to assess for instrumental ADLs. For both measures, deviation from the standardised protocol, with use of fewer or additional question items was commonly observed.
Methodological quality assessment
For articles reporting the prevalence of disability in specific items of ADL, the median (range) methodologic score was 5 (4–6) out of a possible 8. Participant characteristics were generally adequately described, whilst the participation response rate and characteristics of patients declining participation were less frequently reported. A post-hoc decision was made not to use quality criteria in any sensitivity analysis due to the heterogeneity across studies.
ADL disability in adults with cancer
Disability relating to basic ADLs
Eighteen studies provided information about performance of basic ADLs. The mean [95% CI] prevalence of disability was 36.7% [29.8–44.3] for the overall population (see Fig. 2). There was a wide range of prevalence estimates for basic ADL disability across both inpatient (14–53%) and outpatient settings (0–86%) and significant heterogeneity across study-specific estimates (I2 96% for both, p < .001). The pooled mean estimate was not statistically different according to setting (inpatient 32.0% [24.0–41.3] vs. outpatient 41.2% [31.2–51.9], p = .19).
Fig. 2Prevalence of disability in basic and instrumental ADL by study. Red lines indicate inpatient settings and black lines outpatient settings.
Fifteen studies provided information about instrumental ADLs. The mean [95% CI] prevalence of disability was 54.6% [46.5–62.3] for the overall population (see Fig. 2). Again, there was a wide range of prevalence estimates for disability across both inpatient (44–77%) and outpatient settings (13–75%) and significant heterogeneity across study-specific estimates (I2 91 and 97%, both p < .001). The pooled mean estimates of disability according to study setting were borderline statistically different (inpatient 62.8% [55.2–69.8] vs. outpatient 46.1% [31.9–61.0], p = .05).
Number of ADLs affected and ADL disability scores
In the inpatient setting, the number of basic ADLs requiring assistance ranged from 4.7 to 5.0, and instrumental ADLs ranged from 3.0 to 4.7. In the outpatient setting, the number of basic ADLs requiring assistance ranged from 0.6 to 4.0, and instrumental ADLs ranged from 0.9 to 3.1 (Table 2). Seventeen studies provided mean total scores using a standardised instrument, which may be useful to inform future trial design. Study specific estimates from commonly used instruments across different cancer types, stages, and clinical settings are shown in Table 3.
Table 2Characteristics of studies and number of ADL disabilities.
Abbreviations: ADL: Activities of Daily Living; IADL: Instrumental Activities of Daily Living; ns; not stated; OARS: The Older Americans Resource and Services
a Median was reported in study, and assumed as mean.
b Median was reported in study, and assumed as mean. Multiple groups were combined into one group using Cochrane formula.
All values rounded to nearest decimal place. For all outcome measures, except MDS-ADL, a lower score reflects dependence, while a higher score reflects independence.
Abbreviations: ADL: Activities of Daily Living; FAI: Functional Assessment Inventory; FIM: Functional Independence Measure; MDS – ADL: Minimum Data Set Activity of Daily Living; IADL: Instrumental Activities of Daily Living; IFS – CA: Inventory of Functional Status-Cancer; ns; not stated; OECD: Organisation for Economic Co-operation and Development.
a Means combined from multiple samples using Cochrane formula for combing groups.
Among eight basic ADLs assessed, disability relating to personal hygiene was the most prevalent, followed by disability related to walking, transfers, and bathing. In four of the studies on walking, two reported a prevalence of disability ≥ 50%. The least affected activities were eating and stair climbing (Fig. 3). Among eight instrumental ADLs assessed, disability relating to housework was most prevalent, followed by disability related to shopping and transportation. Where performance in housework was assessed, four of six studies found ≥ 50% of patients reporting disability. The least affected activities were telephone use, handling finances, and medication management (Fig. 4).
Fig. 3Prevalence of disability in specific basic ADLs. Red lines indicate inpatient settings and black lines outpatient settings.
This systematic review of 43 observational studies, with 19,246 patients, reveals that about one-third of adults with cancer respectively require assistance to perform basic ADLs and about one-half require assistance to perform instrumental ADLs. Although a wide range of instruments, item and scales were used, our analysis of specific ADLs demonstrates that limitations most frequently related to the basic ADLs of personal hygiene, walking and transfer, and the instrumental ADLs of housework, shopping and transportation. Our findings highlight substantial need for rehabilitation services that focus on maintaining functional independence, and underscore an important role for professionals skilled in occupational assessment and therapy.
Measurement of ADL disability
The most common instruments in measuring ADL disability were the Katz Index of Independence in ADL and the Lawton IADL Scale. Some studies included fewer or additional ADL items using modifications of these standardised measures. A small but significant proportion of studies utilised questionnaire of selected ADL items. The use of selected items to assess ADL performance has been advocated for in clinical practice. Roehrig and colleagues [
] recommended geriatricians to use six ADL items instead of 18 items when conducting screening assessment of frail older people. This recommendation was made based on analysis of ADL data from 327 patients entered into a forward selection model. Six ADL items were sensitive in identifying 98.5% of patients with ADL related disability. Deviations from the use of standardised measures and use of self-constructed questionnaires are stumbling blocks to the advancement of science in the study of disability, as data on ADL disability could not be compared across studies or pooled together to increase power.
Reporting of ADL related disability
All the studies used dependency, defined as having difficulty to do an activity or having someone’s help in at least one ADL, as an indicator of disability. A measure of dependency provides an indicator of disability, but is not a measure of disability itself [
]. Disability is often measured by the degree of difficulty (none, some, a lot, unable) or the level of help needed (minimum assistance, moderate assistance, maximum assistance) with a task. Much information about the severity of disability is therefore lost when a measure of dependency is used. Verbrugge and Jette’s [
] critiqued that whilst dependency measures the need for an intervention to reduce disability, it is not discriminative in capturing the characteristics of disability, and therefore not able to provide patient-specific information about rehabilitation needs or home care support needs.
Characteristics of ADL related disability
Prevalence of ADL disability
ADL related disability was highly prevalent among adults living with cancer. Approximately one-third (37%) and half (55%) of those studied experienced difficultly of required assistance to complete basic and instrumental ADLs respectively. These findings are not unexpected as literature has previously discussed the severe impact of cancer and its treatment on ADL performance. For example, Mohile et al. [
] found 41% and 44% of older community-dwelling people with cancer experienced difficulty with at least one ADL. Given the high incidence of cancer worldwide, and the resultant inability to manage ADLs, management of ADL related disability should arguably be a core part of oncology practice.
There was significant heterogeneity across study-specific prevalence estimates for disability relating to basic and instrumental ADLs. This finding likely relates to differences in the characteristics of each study. A systematic review by Harrison et al. [
] on care needs at different phases of cancer experience found unmet care needs in ADL ranged widely (1–73%). Further analysis showed that care needs for ADLs changed according to the point in the cancer trajectory: at diagnosis (5–10%), during treatment (5–73%), post-treatment (41–47%); and during the terminal stages of illness (1–52%). This heterogeneity sheds new light on the controversy surrounding the impact of cancer and its treatment on ADLs, which warrants further study.
ADL disability higher in inpatients than outpatient
The prevalence of disability related to instrumental ADLs was higher in studies of inpatients compared to outpatients. Similarly, Retornaz et al. [
] found only 23% of similar but community-dwelling group reported disability. One would expect more disability in inpatients as the inability to perform instrumental ADLs at home may be one reason for admission [
]. Such information may be useful to plan and evaluate strategies to prevent disability in outpatient oncology settings.
A hierarchical relationship in ADL
The most commonly affected basic ADL was personal hygiene, such as grooming and brushing teeth, though this finding should be interpreted with caution as the activity was only assessed in one study. Walking and transfers were also commonly affected and these finding corroborate previous studies [
]. Many clinicians would intuitively agree as walking and transfers as mobility tasks are physically demanding. The least affected basic ADL was eating, consistent with the order of loss in ADL functions proposed by Katz and colleagues [
]. In general, instrumental ADLs requiring physical function were more commonly affected than those requiring cognitive functions. This may represent a bias in the selection and recruitment of patients into studies, e.g. excluding those without capacity to consent, but may reflect that patients still perform cognitive tasks despite the loss of physical functions. Engaging patients in cognitive activities to maintain independence in instrumental ADL is encouraged.
Strengths and limitations
There are several strengths to this review. We developed a detailed protocol that ensured consistency and transparency in the review processes. Publication bias was minimised by utilising more than one source of information and methods to locate published and unpublished studies. Our searches were not restricted to a certain discipline and spanned health and social care databases. We also took steps to minimise judgment errors and bias, with authors’ independently extracting and reviewing data. There were challenges in reaching out to experts to identify grey literature and when requesting additional data from the retrieved articles or published abstracts, which could have led to the omission of relevant on-going studies. Due to lack of time, resource (financial and expertise) and facilities for translation, we could only screen and include studies published and written in English language.
Future work
To advance the science of future studies relating to ADL disability, we encourage researchers and clinicians to work towards a consensus on the definition of disability to influence more uniform measurement of ADL disability.
We strongly advocate for the use of standardised, validated instruments over local variations. Use of validated tools which consider ADLs, for example the Lawton IADL Scale or the Functional Independence Measure, would allow data to be reliably compared across future studies and populations. We suggest that subjective measures of dependency used as proxies for disability, or simply summing the number of ADLs that require assistance, does not portray the full nature or extent of ADL related disability. Future research in this field could consider the degree of difficulty, type and level of assistance required to perform specific activities. Finally, intervention studies targeting the maintenance of ADL in people living with cancer, including advanced and progressive disease, should select valid and sensitive scales to help identify, develop and evaluate the best approaches.
Conclusions
This systematic review has identified the nature and prevalence of ADL related disability in adults living with cancer. Overall, about one-third and half of adults with cancer respectively have difficulty or require assistance to perform basic and instrumental ADLs across both inpatient and outpatients settings. Our findings provide useful insights for health care professionals to plan and deliver rehabilitation services, and highlight a particular need for rehabilitation focused on functional independence. The most frequently affected ADLs were personal hygiene, walking and transfers, housework, and shopping and transportation. This understanding can help clinicians focus attention and anticipate problems on activities most commonly affected by cancer. Our findings also underscore the importance of professionals skilled in occupational assessment and therapy in cancer care.
Declaration of interests
MM and IJH received funding from the National Institute for Health Research and from Cicely Saunders International. MM received personal fees from Helsinn and Chugai UK outside the submitted work. JN, LF and GW declare no competing interests. These funders had no role in review design, data analysis, data interpretation, or writing of this report. The corresponding author had full access to articles reporting data used in the review and had final responsibility for the decision to submit for publication.
Competing interests
None of the authors have any competing interests to declare.
Acknowledgements
This work was supported by Cicely Saunders International and the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLARHC) South London at King’s College Hospital NHS Foundation Trust. IJH is an NIHR Emeritus Senior Investigator. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.
Contributors
Substantial contributions to the conception and design of the study: JN, MM; substantial contribution to the acquisition of Data: JN, LF, MM; Analysis and interpretation of the data: JN, GW, LF, IJH, MM; First draft of the manuscript: JN, MM; Revision of the manuscript critically for important intellectual content: All authors; Approval of the final manuscript: All authors; Accountability for all aspects of the work: JN, MM.
(CRD) Cfrad. Systematic reviews: CRD's guidance for undertaking reviews in health care. York (UK): Centre for Reviews and Dissemination, University of York; 2009.
End-of-life care during and after an acute hospitalization in older patients with cancer, end-stage organ failure, or frailty: a sub-analysis of a prospective cohort study.
Comprehensive geriatric assessment adds information to Eastern Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study.
30162-7&id=gr1.jpg){kind=link}
30162-7&id=gr2.jpg){kind=link}
30162-7&id=gr3.jpg){kind=link}
30162-7&id=gr4.jpg){kind=link}