Reduction in uptake of PSA tests following decision aids: systematic review of current aids and their evaluations

https://doi.org/10.1016/j.pec.2004.06.009Get rights and content

Abstract

A man’s decision to have a prostate-specific antigen (PSA) test should be an informed one. We undertook a systematic review to identify and appraise PSA decision aids and evaluations. We searched 15 electronic databases and hand-searched key journals. We also contacted key authors and organisations. All decision aids and evaluations that discussed PSA were included, with meta-analyses performed on two outcomes from the evaluations: PSA testing and patient knowledge of PSA and related issues. Seven decision aids and 11 evaluations were included. The meta-analysis showed a significantly reduced probability in PSA testing after a decision aid: −3.5% (95% confidence interval: 0.0 to 7.2%; P = 0.050). There were significant improvements in knowledge within 2 weeks after a decision aid: 19.5% (95% confidence interval: 14.2 to 24.8%; P < 0.001). The effect on knowledge was less pronounced within 12–18 months after a decision aid: 3.4% (95% confidence interval: −0.7 to 7.4%; P = 0.10). PSA decision aids improve knowledge about PSA testing, at least in the short term. Men given these decision aids seem to be less likely to have the PSA test.

Section snippets

Background

Controversy surrounds the use of prostate-specific antigen (PSA), a serum tumour marker that is widely used as a test to detect prostate cancer. The difficulties lie in the poor sensitivity and specificity of the test and our lack of understanding of both the natural history of prostate cancer and of how best to treat the disease [1], [2]. Internationally, the situation varies: in the USA, for instance, there are well-established PSA screening programmes [3]. In the UK, the government has

Objectives

We undertook a systematic review to identify and appraise PSA decision aids and evaluations of decision aids.

Criteria for considering studies for this review

We included all identified PSA decision aids and the evaluations (quality and outcomes) of PSA decision aids. We excluded prostate cancer/screening decision aids which did not discuss PSA testing and prostate cancer treatment decision aids. In addition, we excluded decision aids and evaluations involving patients and prisoners in military settings — where they may not be free to choose [7]. We did not exclude articles or aids on design or evaluation quality grounds: these were assessed

Search strategy for identification of studies

Fifteen electronic databases were searched: [MEDLINE (1966–2003), EMBASE (1980–2003), CINAHL (1982–2003), CancerLit (1983–2003), Cochrane Database of Systematic Reviews (2003), Cochrane Central Register of Controlled Trials (2003), Science Citation Index (SCI) (1981–2003), Social Sciences Citation Index (1981–2003), Applied Social Sciences Index and Abstracts (1987–2003), NHS National Research Register (2003), Psycinfo (1967–2003), Educational Resources Information Centre (ERIC) (1985–2003),

Methods of the review

The titles and abstracts of all identified articles were assessed by one reviewer (RE) for their inclusion/exclusion. A random sample (10%) of the included and excluded articles was checked by two independent reviewers (AE, JB: 5% each). A list of articles to be assessed in full-text was then agreed upon. One reviewer (RE) assessed all the full-text articles for inclusion. They were also distributed between three other reviewers (AE, JB, EW) for independent assessment. A list of articles for

Description of studies

Thirty-two thousand five hundred and ninety-six articles were identified in the electronic search, from which 35 articles were assessed in full-text for possible inclusion. The yield from the individual databases is shown in Appendix B. Hand-searching and individual contacts identified a further eight articles. A total of 43 articles were therefore identified for assessment in full-text, from which emerged a final ‘inclusion’ list of 7 decision aids and 11 decision aid evaluations (Table 1,

Methodological quality of included evaluations

The methods score of the evaluations is shown in Table 2. The studies were scored independently by two reviewers and both scores are shown. All studies were scored out of 22, the exception being the observational study by Beggs et al. [27] which was scored out of 16. In comparison to studies generally encountered in this research field [7], the trials by Wilt et al. [24], Volk et al. [20], Wolf et al. [17] are above average in their methodological quality; the trials by Schapira and VanRuiswyk

PSA testing

In the four randomised controlled trials [21], [24], [25], [26], PSA testing was assessed at 12 months after the intervention, with the exception of Hammond et al. (18 months) [26]. Overall, there was a 3.5% absolute reduction in the number of patients who had a PSA test (95% confidence interval: 0.0–7.2%; P = 0.05) following a decision aid intervention. The results are presented in Table 3 and Fig. 1. The Hammond et al.’s trial dominated the meta-analysis, but adjusting for the cluster effect

Discussion

Seven PSA decision aids and 11 evaluations were identified. The meta-analyses suggest that the decision aids increase users’ knowledge about PSA and reduce uptake of the test. Most of the decision aids were of a paper/document format [13], [14], [15], [17], [18], [19] and the majority were from the USA [13], [14], [15], [16], [17]. Whilst there was development information available on all seven decision aids, only three had been formally evaluated [14], [16], [17]. All of the evaluations were

Acknowledgements

We thank Tricia Chapman, Library Services, Ceredigion and Mid Wales NHS Trust and Lori Havard, Library Services, University of Wales Swansea, for acquiring the papers. We thank Professors Annette O’Connor, Hanneke De Haes, Pat Wright and Dr. Bob Volk for sharing information and for their kind advice.

Contributions: RE designed the protocol, conducted the literature searches, assessed studies for inclusion and data extraction, and led the drafting of this article. He is also the guarantor of the

References (36)

  • R. Jepson et al.

    The determinants of screening uptake and interventions for increasing uptake: a systematic review

    Health Technol. Assess.

    (2000)
  • E.J. Matthews et al.

    Efficient literature searching in diffuse topics: lessons from a systematic review of research on communicating risk to patients in primary care

    Health Libr. Rev.

    (1999)
  • A.M. O’Connor et al.

    Decision aids for people facing health treatment or screening decisions

    (2001)
  • Edwards A, Unigwe S, Elwyn G, Hood K. Personalised risk communication in health screening programs. Cochrane Database...
  • J.J. Deeks et al.

    Statistical methods for examining heterogeneity and combining several results in meta-analysis

  • Cancer Information Services

    Questions and answers about the prostate-specific antigen (PSA) test

    (2001)
  • VA Medical Centre Minneapolis

    Early prostate cancer information leaflet

    (2001)
  • American Institute for Cancer Research

    Reducing your risk of prostate cancer

    (1999)
  • Cited by (65)

    • Effectiveness of interprofessional shared decision-making training: A mixed-method study

      2022, Patient Education and Counseling
      Citation Excerpt :

      SDM aims to provide individualized care and incorporate the personal context by fostering patient engagement in decisions concerning their health [2–4]. SDM can improve the clinical decision-making process by reducing the overuse of options not associated with benefits [5] and by enhancing the use of options associated with benefits [6]. This strategy is considered the pinnacle of patient-centered care [7].

    • Patients’ perceptions and attitudes on recurrent prostate cancer and hormone therapy: Qualitative comparison between decision-aid and control groups

      2017, Journal of Geriatric Oncology
      Citation Excerpt :

      One way anxiety is reduced is through improved knowledge, in turn reducing uncertainty [9–12]. Several studies have shown, DAs are effective for increasing patient knowledge [5,13–15] through providing an improved understanding of specific mechanisms implicated, such as increasing knowledge of treatment options, their advantages and disadvantages [3] and, reducing anxiety [16]. Also, DAs have several advantages such as: increasing patients' participation in care-decisions, resulting in decreased use of PSA testing [17]; facilitating shared decision-making between patient and physician [13,18,19]; lowering patient levels of decisional conflict [19]; and eliminating searches for additional information [1].

    • Development of PRIDe: A tool to assess physicians' preference of role in clinical decision making

      2012, Patient Education and Counseling
      Citation Excerpt :

      Drawing its content from current clinical practice guidelines [24–28] and SDM principles [2], DECISION+ consists of three interactive workshops of three hours each, paper reminders on key issues highlighted in the workshops delivered at the point of care, and feedback on the agreement between patients’ and providers’ level of decisional comfort [23]. The workshops are based on case scenarios and cover [1] the probabilistic nature of a bacterial vs. a viral acute respiratory infection, [2] evidence of the balance between risks and benefits in the use of antibiotics to treat acute respiratory infections, [3] strategies to communicate this information to the patient, and [4] strategies to foster the patient's active participation in decision making. The workshops use videos and reflective exercises to facilitate group discussion.

    View all citing articles on Scopus
    View full text