1.1 Distress in cancer patients and outcomes of nontreatment
Cancer patients often experience high levels of psychological distress which can have a negative effect on their quality of life and treatment outcomes.1 Recent studies including systematic reviews have demonstrated that up to half of individuals living with cancer report clinically significant levels of distress.2–6 Clinically significant levels of distress is an indicator of conditions, such as depression or anxiety, which require further clinical intervention. Distress levels are higher in cancer patients than that of the broader society, yet often go unrecognized and untreated by health professionals.2,7,8
The National Comprehensive Care Network (NCCN) developed guidelines detailing standards of care for distress screening.1 The NCCN guidelines define distress as “a multifactorial unpleasant experience of a psychological, social, or spiritual manner that may interfere with the ability to cope effectively with the symptoms, treatment, and overall experience of cancer.”1 Unaddressed distress is associated with intensified physical symptoms, social disruptions, and negative effects on treatment adherence.9,10 As well as reduced quality of life, untreated psychological distress can lead to increased risk of mortality.5 Overall, there is substantial evidence demonstrating poorer outcomes for patients experiencing distress while managing a cancer diagnosis and treatment.1,2,5,9,10 Clinicians report lack of time and resources as a barrier to implementing a comprehensive evidence-based screening pathway in their service,11 demonstrating a need for further research regarding the characteristics of services which facilitate routine distress screening and management.
1.2 Guidelines and quality screening
According to international guidelines, distress should be identified, monitored, and treated throughout all stages of the disease and in all settings.1 Although it is beneficial for health professionals to be alert to signs of distress, Australian guidelines recommend a formalized screening pathway using validated tools to measure presence of distress in cancer patients.12 This is similar to other international guidelines and ideally screening pathways include a 2-step assessment process, using first a short screening tool, followed by a more comprehensive tool or assessment for those with a positive initial screen.12 Referral networks and processes are paramount to quality distress screening and management and are featured prominently in guidelines.1,12 Without appropriate referral pathways, screening alone is unlikely to have positive impacts on patient wellbeing.13
1.3 Service characteristics associated with service-wide process
Numerous studies have been conducted to examine cancer services’ adherence to screening pathways within the United States.2,14,15 For example, a review of cancer services demonstrated that 62% of services complied, with the highest rates (76.3%) being observed in Cancer Community programs, and the lowest (43.3%) within National Cancer Institute designated Comprehensive Cancer Center Program.14 However, these studies did not describe specific components of a screening pathway that had been implemented and focus specifically on cancer services within the United States healthcare system. Furthermore, there are limited studies that link specific service characteristics with providing high-quality service-wide distress screening and management. Such data could provide opportunity to inform efforts to achieve widespread and equitable access to comprehensive supportive care.
The Australian health care system is a hybrid model whereby public insurance for basic services is provided by the government (eg, Medicare) but can be supplemented if individuals purchase private insurance for additional health care coverage.16 This model is similar to that functioning in Belgium, Canada, and France. In Australia, cancer care can be delivered via public or private health services and providers. Although some studies have addressed economic barriers to accessing mental health facilities,2 a link between high-quality distress screening through adherence to guidelines and service characteristics has not yet been explored in Australia.
This study aimed to characterize Australian cancer services’ distress screening and management processes as reported by service representatives. A cross-sectional survey study was conducted to address the following:
- 1) What proportion of representatives report the presence of a service-wide distress screening and management process and partial implementation of distress screening compared to no screening procedures?
- 2) What service characteristics are significantly associated with presence of screening process compared to those who do not screen?
In answering these questions, this study reflects on the progress made toward achieving comprehensive supportive care in Australian cancer services. Given the International Psycho-Oncology Society's emphasis on the identification and management of distress in all settings, cancer services in this study broadly refer to any tertiary setting that provided medical treatment or support for cancer patients.17 This could include medical oncology, radiation oncology, surgical oncology, in-patient cancer services, intravenous chemotherapy infusion centers, psycho-oncology, and palliative care services.
3.1 Study design and setting
This study was part of a cross-sectional survey of Australian cancer services’ utilization of distress screening and management guidelines and preferred implementation strategies.18 The project was approved by the Hunter New England Human Research Ethics Committee. Data collection occurred over an 8-month period from January 2017 to August 2017. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.19
3.2.1 Cancer service identification
There is no centralized database or directory of cancer services in Australia. Therefore, a comprehensive list of Australian cancer services was constructed using multiple publicly-available online sources such as CanRefer (an online database administered by Cancer Institute New South Wales) and hospital directories or networks.20 A member of the research team contacted the health service directly if it was not clear from online sources if the service provided care for cancer patients (eg, out-of-date website). A total of 240 health services were identified as providing some form of cancer care. Larger hospitals often contained >2 cancer-specific health unit, for example, a medical oncology service and a radiation oncology service. In this case, it was counted as 2 health services and 2 representatives were approached to provide details specific to a single service only.
3.2.2 Participant recruitment and eligibility
Clinical leads or service managers of identified health services received a personalized email. A brief study description was provided, along with a request to nominate potential individuals within their team who could provide detailed information on screening practices and management. Only one representative per cancer health service was contacted. Reminder emails and phone calls were used to prompt service representatives who did not respond to initial emails.
Nominated individuals received a personalized email informing them that they had been identified as a potential service representative. The email contained a brief study introduction and an embedded survey link. Nominated individuals who did not respond to the initial email received email prompts and telephone reminders.
From the 240 services identified, a total of 122 (51% response rate) representatives were recruited with 110 representatives providing complete data (90%) for the analysis. Representatives were asked at the start of the survey to confirm that the service provided care for cancer patients and indicate if they were responding on behalf of: medical oncology, radiation oncology, surgical oncology, in-patient cancer service, intravenous chemotherapy infusion center, psycho-oncology, and palliative care. Individuals who did not confirm this or did not select any of the above did not complete the survey. There was no eligibility criteria related to the type or stage of cancer which services may treat. Of the 122 representatives, 39 respondents provided details on a unique clinic in a multiservice site (ie, medical oncology, radiation oncology).
3.3 Study Measures
Three survey sections were developed for the study:
3.3.1 Distress screening and management pathway
This section focused on distress screening practices and guideline implementation. The items were adapted using a previous NCCN audit on integrating psychosocial care into routine practice.21 Respondents were asked whether patients were screened for distress or emotional wellbeing, including whether all patients were screened or only specific patient groups, how distress was identified, if a referral protocol was in place, and if patients were rescreened.
Three distinct groups were defined according to their responses: an implementation of evidence-based pathway group; a partial implementation group; and a nonscreening group. An evidence-based screening pathway was defined as a process where all outpatients were screened; a validated tool was used for screening with follow-up assessment as required; with referrals generated by a health professional in accordance with established protocols with rescreening. All 3 conditions must be met. If only 1 or 2 of the conditions were met, the service was classified as partial implementation (Supplementary Table, http://links.lww.com/OR9/A6). The nonscreening group were services that did not screen patients at all.
3.3.2 Service characteristics
Respondents reported: state-based location; funding structure (eg, privately funded only); regional setting (defined as centers with a population of <100,000); and availability of mental health services.
3.4 Statistical methods
Descriptive analyses were used to report survey data. Missing data were minimal throughout the study (90% full completion); only participants who completed all service characteristic and screening items were included in analyses. Univariate χ2 tests were used to identify significant associations between distress screening implementation and service characteristics. A P value of ≤.05 was considered statistically significant. Data analyses were completed with Stata 11 software (StataCorp, College Station, TX).
4.1 Service characteristics
A total of 122 (51% response rate) representatives responded with 110 representatives providing complete data (90%). Characteristics of the 110 services represented are listed in Table 1. Briefly, the majority of representatives responded on behalf of publicly funded services (68%) and were based in urban settings (66%). Over half (65%) were identified as being affiliated with a teaching or university hospital. Four representatives indicated that they were unsure if they were located in a rural or urban setting. All Australian states and territories were represented, with the majority from the 3 most populous states: New South Wales (35%), Queensland (23%), and Victoria (20%). Approximately 4 in 5 services (78%) had on-site mental health services. Individuals reported holding single roles as: nurse unit managers (37%); oncology-specific nurses (24%); cancer care coordinators (15%); social workers (8%); service directors or administrators (7%); psychologists (5%); or general nurses (4%). One in 5 respondents (22%) held >1 role with the most common combination being an oncology-specific nurse and cancer care coordinator. Four physicians participated (4%). Approximately 83% reported a personal role in screening or managing patients’ emotional needs. Of the 17% who reported no personal role, the majority were service directors or nurse unit managers.
4.2 Distress Screening and Management Practices
Of the participating services, 22 (20%) representatives reported that the service did not screen for distress at all and 80 (72%) reported the service had partial implementation of distress screening and management pathways. The majority of respondents reported interest in implementing or improving an evidence-based pathway within nonscreening or partial implementation services (75% and 76%, respectively).
Only 8 services (7%) had implemented an evidence-based pathway comprised of routine distress screening, referral pathways, and rescreening. Of these 8 services, 3 (38%) had participated in a research program and 6 (75%) had evaluated the pathway effectiveness and uptake. In 5 of these services, the responsibility for screening was allocated to a team (ie, >1 health professional) including care coordinators, nurses, and social workers. In 2 services, it was allocated to nurses alone and in 1 service to care coordinators alone. Only 1 service included physicians as being individuals responsible for screening and only as part of a wider team.
4.3 Service characteristics associated with service-wide process
To highlight inequitable access to any form of distress screening and management, services classified as complete or partial implementation were grouped and compared to services that did not complete screening in any form (Table 2). Two of the 4 service characteristics analyzed were significantly associated (P ≤ .05) with distress screening implementation: funding structure (ie, public funding) and affiliation with university (ie, teaching hospital). There was no association reported between location (rural or urban) or on-site mental health services.
Internationally, there is growing recognition that distress screening is not being consistently offered in cancer services and where screening is occurring, referral rates and uptake for distressed patients remain low.14,22 These inconsistencies represent missed opportunities to improve quality of life, treatment adherence, and mortality rates in cancer patients,1,2,5,9,10 and have the potential to result in inequitable access to supportive care services.
This study categorized the distress screening practices as reported by 110 service representatives and explored potential associations with service characteristics. The resulting data are not specific to any particular cancer type or stage and therefore summarizes the practices of a wide variety of services. Given the IPOS emphasis on universal screening in all settings, this is the strength of this inclusive dataset. These data can determine areas whereby patients may not be screened for distress or provided appropriate supportive care referrals, and highlight specific types of services that may require additional support to integrate this evidence-informed practice into routine care.
In summary, the study found only a small proportion (7%) of Australian cancer service representatives reported the presence of a complete distress screening and management pathway. Progress is being made, however, with the majority of services considered as partial implementation. Although this is promising, 20% of services report not screening at all.
5.1 Pin-pointing the gap in evidence-based distress screening and management processes
This study found 2 potential service characteristics were associated with the level of evidence-based distress screening implemented throughout health services: a public funding structure and being a teaching hospital. No significant difference was found between rural and urban services or the presence of mental health services.
5.1.1 A smaller proportion of privately funded health services follow consistent screening and referral practices
The survey found that services receiving public funding reported complete or partial implementation of screening processes (88%) at a significantly higher proportion than their privately funded counter parts (63%). Providing consistent screening practices across private and public cancer services is paramount to quality patient-centered care, particularly given that many Australian cancer patients and patients in similar health systems with hybrid funding models (eg, Canada, Belgium, and France) rely on private health insurance to supplement government-subsidised services.
The provision and quality of supportive care services in publicly funded services is routinely audited as part of Australian government reporting.23 However, this quality assurance process may not be undertaken in private services and there is limited information on the provision or quality of supportive care services outside of public services. There is conflicting information on how patients and health professionals might perceive differences in the quality of care according to funding source. For example, an Australian survey of patients with hematological cancer reported that the use of the public health care system was associated higher perceived quality of care and follow-up care by patients.24 In addition to this, a study of health professionals’ perceptions in rural Australia suggests that workers believe that patients who receive care in private settings are less likely to be referred for support services in their community.25 However, a 2008 study indicated that although Australians have greater trust in the public system (Medicare) than private system, they hold greater trust in private compared to public hospitals.26 This suggests that some Australian cancer patients may not be aware of the potential gaps in private services (such as the gap in distress screening and referrals found in this study) before opting out of public services.
The discrepancy in accessing high-quality supportive care services according to funding structure is relevant to other hybrid private/public health systems operating internationally, and efforts to mitigate inequitable access will require coordinated health policy and evaluation at multiple levels. Using Australia as an example,27 the Australian Department of Health released the “Public and Private Cancer Care Project” report as part of the National Cancer Work plan, to address the enablers and barriers to effective cancer care across private and public services.28 According to the report, variability in practices across public and private sectors limits the ability to implement service improvements, including distress screening.28 The report highlights that a standard of care should be implemented between the 2 settings through agreed pathways of cancer care. However, the findings suggest that public hospitals were perceived to provide a broader and more comprehensive level of supportive care, with private hospitals being considered “less coordinated” for patients, and that fees impacted access to psychosocial care.28 This is consistent with the results of our study, in that public hospitals are more likely to be providing evidence-based screening practices within their service than private hospitals, indicating a variation between the differently funded hospitals.29
5.1.2 Services not affiliated with a university or teaching hospital
A significantly larger proportion of health services affiliated with a university had some form of distress screening procedures (88%) as compared to health services without an affiliation (64%). This reflects the health care system's broader challenge of translating evidence-based guidelines into policy and practice,30 in that a close connection with research and teaching can help to drive evidence-based care. Cancer services and the field of psycho-oncology have a responsibility to acknowledge these challenges and push to expand evidence-based screening and management to health services outside of academic centers and ensure equitable access to supportive care.17
Service characteristics are not associated with implementation of evidence-based distress screening and management processes
5.1.3 Availability of on-site psychosocial care
Previous studies report clinicians perceive a lack of psychosocial support services as a barrier to implementing distress screening.31 However, this study found no significant association between availability of mental health services, such as psychologists, psychiatrists, or social workers, and the implementation of distress screening and management. This is in line with a 2015 cross-sectional study in which the size of facility, location, available mental health services, and cancer center type had little bearing on whether or not services had begun implementing distress screening.15 These findings must be cautiously interpreted as the relatively small sample sizes in both studies (70 and 110 services) may not have allowed for sufficient power to detect a difference.
15Clinicians’ perception of accessibility of psychosocial services (eg, availability, waiting times, or cost)32,33 or clinicians’ clarity on referral pathways,34,35 is an important consideration in the implementation of evidenced-based screening practices. The simple existence of on-site psychosocial services is not sufficient to guarantee appropriate referrals by individual clinicians. Efforts to implement evidence-based screening and management in an Australian setting should focus not just on health system's capacity to service referrals but also on education and training, promotion of supportive care services, and improved referral pathways.17,33
5.1.4 Regional health services
Our study found that there were no significant differences between rural and urban services regarding distress screening and management implementation. Although some studies have found no evidence indicating that increasing remoteness is associated with higher levels of distress in Australian cancer patients, a 2012 systematic review of psychosocial wellbeing in rural and urban areas identified 4 studies reporting poorer outcomes for rural patients.36–38 The discrepancy in outcomes suggests that a more comprehensive investigation is warranted when understanding the psychosocial needs for rural patients. Although distress screening may be being implemented at a consistent level between area groups, Australian cancer patients that live rurally have needs that are unique and separate from their urban counterparts.39 This includes geographical limitations when accessing psychosocial services,36 higher levels of stigma surrounding mental health support, isolation from support groups, and cost of travel to necessary services.36,37 It is important these needs are considered when evaluating implementation strategies to address the gap between supportive need outcomes in rural and urban patients.
Although this study has uncovered some of the discrepancies in cancer care between health service characteristics, a more in-depth analysis is required. The availability and number of service characteristics for this analysis will be limited by sample size and reporting. Response bias and incompletion rates can be high in online surveys, posing a limitation in studies that use these data. This study achieved a 51% response rate, and 10% of respondents did not provide sufficient data to be included in the analysis (eg, screening or service information). We cannot confidently compare the characteristics of nonresponding service representatives to respondents given there is limited or inconsistent information available which describes Australian cancer services. A recent systematic review of survey studies with health professionals reported an average response rate (38%) for online surveys.40 Given the comparatively high response rate (51%) and the low incidence of self-reported evidence-based practice, response bias within this study is likely to be minimal. It is also important to note that service representatives were asked to report service-wide practices. Although the survey did ask individuals to note any differences between personal practice and service procedures, representatives may not have been privy to all individuals’ screening practices within the service.
As the scope of this study is to characterize Australian cancer services, rather than the communities these services serve, this article is only able to speak to equitable access to evidenced-based psychosocial care for Australians based on a few service characteristics, for example, rural versus urban areas. Further research should address service provision of equitable access to evidence-based psychosocial care in health settings serving Aboriginal, culturally and linguistically diverse, and lower socioeconomic communities.
The results of this study indicate that 80% of service representatives report that their cancer service was providing complete or partial implementation of evidence-based psychosocial distress screening and management. This suggests services are able to provide best-practice supportive care to cancer patients. It is concerning, however, that 20% of cancer services are providing no distress screening to their patients at all. Given Australian patients’ high reliance on private health care, and the fact that services that are providing complete or partial implementation of screening practices are associated with characteristics such as being a publicly funded or a teaching hospital creates concerns for equitable access to supportive care. Although whether services reported that their service was located in an urban or regional area, or had on-site psychosocial services was not significantly associated with provision of complete or partial implementation practices, this does not mean that these characteristics to do not have an impact on implementation. Further research should consider how to best help private services overcome barriers to evidence-based screening practices and the impact of public private partnerships on rural and regional cancer patients and begin working toward a better understanding of clinician psychosocial referral pathways.
Conflicts of Intererst
The authors have no conflicts of interest to disclose.
. National Comprehensive Cancer NetworkNCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw
. Jacobsen PB, Ransom S. Implementation of NCCN distress management guidelines by member institutions. J Natl Compr Canc Netw
. Reilly CM, Bruner DW, Mitchell SA, et al. A literature synthesis of symptom prevalence and severity in persons receiving active cancer treatment. Support Care Cancer
. Skarstein J, Aass N, Fosså SD, Skovlund E, Dahl AA. Anxiety and depression in cancer patients: relation between the Hospital Anxiety and Depression Scale and the European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire. J Psychosomat Res
. Russ TC, Stamatakis E, Hamer M, Starr JM, Kivimäki M, Batty GD. Association between psychological distress and mortality: individual participant pooled analysis of 10 prospective cohort studies. BMJ
. Carlson LE, Waller A, Groff SL, Bultz BD. Screening for distress, the sixth vital sign, in lung cancer patients: effects on pain, fatigue, and common problems--secondary outcomes of a randomized controlled trial. Psychooncology
. Gouveia L, Lelorain S, Brédart A, et al. Oncologists’ perception of depressive symptoms in patients with advanced cancer: accuracy and relational correlates. BMC Psychol
. Werner A, Stenner C, Schuz J. Patient versus clinician symptom reporting: how accurate is the detection of distress in the oncologic after-care? Psychooncology
. Lazenby M. The international endorsement of US distress screening and psychosocial guidelines in oncology: a model for dissemination. J Natl Compr Canc Netw
. DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Archives of internal medicine
. Rankin NM, McGregor D, Stone E, et al. Evidence-practice gaps in lung cancer: a scoping review. Eur J Cancer Care
. Butow P, Price MA, Shaw JM, et al. Clinical pathway for the screening, assessment and management of anxiety and depression in adult cancer patients: Australian guidelines. Psychooncology
. Skaczkowski G, Sanderson P, Shand M, Byrne A, Wilson C. Factors associated with referral offer and acceptance following supportive care
problem identification in a comprehensive cancer service. Eur J Cancer Care
. Zebrack B, Kayser K, Bybee D, et al. A practice-based evaluation of distress screening protocol adherence and medical service utilization. J Natl Compr Canc Netw
. Lazenby M, Ercolano E, Grant M, Holland JC, Jacobsen PB, McCorkle R. Supporting commission on cancer–mandated psychosocial distress screening with implementation strategies. J Oncol Pract
. Dixit SK, Sambasivan M. A review of the Australian healthcare system: a policy perspective. SAGE Open Med
. Fradgley EA, Bultz BD, Kelly BJ, Loscalzo MJ, Grassi L, Sitaram B. Progress toward integrating Distress as the Sixth Vital Sign: a global snapshot of triumphs and tribulations in precision supportive care
. J Psychosoc Oncol Res Pract
. Fradgley EA, Byrnes E, McCarter K, et al. A cross-sectional audit of current practices and areas for improvement of distress screening and management in Australian cancer services: is there a will and a way to improve? Support Care Cancer
. Elm Ev, Altman DG, Egger M, Pocock SJ, G⊘tzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ
. Cancer Institute NSW. CanRefer. Available at: https://www.canrefer.org.au/
. Published 2019. Accessed.
. Jacobsen PB, Lee M. Integrating psychosocial care into routine cancer care. Cancer Control
. Zebrack B, Kayser K, Sundstrom L, et al. Psychosocial distress screening implementation in cancer care: an analysis of adherence, responsiveness, and acceptability. J Clin Oncol
. Information BoH. Snapshot Report: NSW Patient Survey Program. NSW Australia 2014.
. Tzelepis F, Hanna JH, Paul CL, Boyes AW, Carey ML, Regan T. Quality of patient-centred care: medical oncology patients’ perceptions and characteristics associated with quality of care. Psychooncology
. Lee J. Health workers’ perceptions of psychosocial support services for cancer patients in rural Victoria. Cancer Forum
. Hardie EA, Critchley CR. Public perceptions of Australia's doctors, hospitals and health care systems. Med J Aust
. Cancer Institute NSW. NSW Cancer Plan: state-wide plan for lessening the impact of cancers in NSW. https://www.cancer.nsw.gov.au/getattachment/cancer-plan/CancerPlan2016_Web.pdf?lang=en-AU
. Published April 2016. Accessed March 9, 2020.
. National Cancer Care Reference Group (NCERG). Public and Private Cancer Care Project: Final Report. https://www1.health.gov.au/internet/main/publishing.nsf/content/9ADA98F27F90251BCA257DC7007763D3/$File/Public%20&%20Private%20Cancer%20Care%20Project.pdf
. Published January 22, 2015. Accessed March 9, 2020.
. Carey M, Sanson-Fisher R, Clinton-McHarg T, et al. Examining variation across treatment clinics in cancer patients’ psychological outcomes: results of a cross sectional survey. Support Care Cancer
. Rabin B, Glasgow RE. An implementation science perspective on psychological science and cancer: What is known and opportunities for research, policy, and practice. Am Psychol
. Rankin NM, Butow PN, Thida T, et al. Everybody wants it done but nobody wants to do it: an exploration of the barrier and enablers of critical components towards creating a clinical pathway for anxiety and depression in cancer. BMC Health Serv Res
. Abrahamson K, Durham M, Norton K, Doebbeling BN, Doebbeling CC, Anderson JG. Provision of psychosocial care for cancer patients: service delivery in urban and rural settings. J Prim Care Community Health
. Fennell KM, Bamford L, Olver I, Wilson CJ. Good training, systems and funding, not good luck: what hematologists and oncologists believe would make it easier for them to refer their cancer patients to psychosocial care. Transl Behav Med
. Absolom K, Holch P, Pini S, et al. The detection and management of emotional distress in cancer patients: the views of health-care professionals. Psychooncology
. Biddle L, Paramasivan S, Harris S, Campbell R, Brennan J, Hollingworth W. Patients’ and clinicians’ experiences of holistic needs assessment using a cancer distress thermometer and problem list: a qualitative study. Eur J Oncol Nurs
. Butow PN, Phillips F, Schweder J, et al. Psychosocial well-being and supportive care
needs of cancer patients living in urban and rural/regional areas: a systematic review. Support Care Cancer
. Watts KJ, Good LH, McKiernan S, et al. “Undressing” distress among cancer patients living in urban, regional, and remote locations in Western Australia. Supportive Care Cancer
. Murray G, Judd F, Jackson H, et al. Rurality and mental health: the role of accessibility. Australian N Z J Psychiatry
. Gunn K, Turnbull D, McWha JL, Davies M, Olver I. Psychosocial service use: a qualitative exploration from the perspective of rural Australian cancer patients. Support Care Cancer
. Cho YI, Johnson TP, Vangeest JB. Enhancing surveys of health care professionals: a meta-analysis of techniques to improve response. Eval Health Prof