Skip to content

Investigating the true effect of psychological variables measured prior to arthroplastic surgery on post-surgical outcomes: a p-curve analysis

Research output: Contribution to journalArticlepeer-review

Patients’ pre-surgical psychological profiles have been posited to predict pain and function following arthroplastic surgery of the hip and knee. Nevertheless, findings are conflicting, and this may be rooted in biased reporting that makes the determination of evidential value difficult. This ambiguity may have negative consequences for researchers and governmental agencies, as these rely on findings to accurately reflect reality. P-curve analyses were used to establish the presence of evidential value and selective reporting in a sample of studies examining the effect of pre-surgical psychological predictors on outcomes following knee and hip arthroplastic surgery. A systematic search of the literature revealed 26 relevant studies. The examined sets of studies indicate that there is evidential value for the effect of depression on pain intensity and function, anxiety on pain intensity and function, pain catastrophizing on pain intensity, as well as the combined effects of all psychological predictors on pain intensity and function. The presence of evidential value was inconclusive for the effect of optimism on pain intensity. There were no signs that any results were influenced by biased reporting. The results highlight the importance of patients’ psychological profiles in predicting surgical outcomes, which represent a promising avenue for future treatment approaches.

Perspective: The effects of p-hacking are difficult to detect and might be at the root of conflicting findings pertaining to the predictive properties of pre-surgical psychological variables on post-surgical outcomes. P-curve analysis allows the determination of evidential value underlying these relationships and detection of p-hacking to ensure that findings are not the result of selective reporting.
Original languageEnglish
JournalThe Journal of Pain
Early online date21 Oct 2020
DOIs
Publication statusEarly online - 21 Oct 2020

Documents

  • Post print

    Accepted author manuscript (Post-print), 4.26 MB, PDF document

    Due to publisher’s copyright restrictions, this document is not freely available to download from this website until: 21/10/21

    Licence: CC BY-NC-ND

Related information

Relations Get citation (various referencing formats)

ID: 23049204