Écrire un commentaire

This narrative review asks how psychosexual rehabilitation after cancer can be delivered at scale, rather than describing how common sexual dysfunction is. The author organizes the intervention evidence by intensity of care. Low-intensity measures (structured discussion, lubricants and moisturizers, topical lidocaine, brief clinician training) are cheap and reasonably supported. Two widely recommended physical interventions have weaker evidence than assumed: scheduled PDE5 inhibitor "penile rehabilitation" and routine vaginal dilation during radiotherapy. Psychological interventions show small pooled effects, but moderate effects in targeted trials, most clearly therapist-guided internet-based CBT in breast cancer. From this the author proposes a five-step stepped-care model (Steps 0–4, built on extended PLISSIT) with explicit criteria for stepping up and down, plus adaptations for AYA, men on ADT, and sexual and gender minority survivors.

The paper moves the field forward by reframing the problem as one of delivery, by challenging two entrenched practices, and by turning scattered evidence into a model that oncology teams, including nurses, could adopt and test.

Major issues

• Methods are not reproducible. The review relies on "targeted searches" with no databases, dates, search terms, or selection process described. Evidence is "appraised qualitatively" with no tool such as GRADE or risk-of-bias assessment. Claims like "strongest trial evidence in breast cancer" may reflect selective sampling.

• The stepped-care model is not empirically validated. It was derived by the author "mapping" interventions to levels of care, with no consensus process, stakeholder input, or implementation data. The 8–12 week step-up threshold has no cited support. The conclusion that evidence "supports matching intensity to need" goes beyond what the cited studies tested, since none compared stepped care with usual care or direct referral (the paper admits this only at the end).

• Evidence for Steps 0–2 is mostly clinician-level, not patient-level. The cited training studies show clinicians discuss sexuality more often, not that patient outcomes improve. Yet these steps form the base of the model. Feasibility of nurse- or clinician-delivered care also rests largely on one HSCT trial.

• Key claims rest on weak or non-peer-reviewed sources. The HSCT multimodal trial [34] is a conference abstract that is called "important for the model." Other supports [28, 47] are also abstracts. Pilot RCTs (23 and 29 couples) underpin the "medium-to-large" couple therapy claims.

• The paper applies evidence standards inconsistently. It criticizes penile rehabilitation and during-radiotherapy dilation for weak evidence, then recommends supported dilation after radiotherapy based on observational data and an uncontrolled 53-woman cohort (11.5% to 60.9% pre-post), and cites a regimen from an abstract [28].

• Inference from pooled versus targeted effects is speculative. The difference (SMD 0.24 vs. d 0.43–0.72) is taken to show that interventions work best when targeted, which supports stepped care. It could equally reflect heterogeneity, study quality, outcome measures, or small-study effects. Effect sizes of different metrics are also compared without confidence intervals.

• There is no limitations section. It should cover search bias, the breast-cancer skew, scarce data for men, SGM, and LMIC settings, and the lack of cost and workforce evidence.

Minor issues

• PLISSIT is never spelled out, and AYA HOPE is not defined.

• "Psychosexual rehabilitation" is defined as a staged process but used loosely to include medical and physical treatments. A tighter definition would help.

• Table 2 level-of-care labels are confusing ("Steps 0–2", "Step 2 (on-demand use)" for a scheduled intervention). The HSCT row gives no sample size.

• Table 2 says the CBT gains were maintained at 9 months, while the text says 3 and 9 months.

• Table 1's "Rehabilitation targets" column mixes interventions with targets.

• In Table 3, "criteria for stepping up" for Step 0 isn't really a criterion. Criteria such as "relationship breakdown" and "complexity exceeding competence" need operational definitions.

• Figure 1 text is very small and hard to read on a phone.

• The statement that sexual problems are "a leading unmet supportive care need" cites a review limited to men with genitourinary cancers [4].

• The SGM and AYA adaptations are brief and mostly descriptive. Concrete examples of inclusive screening wording would help.

• No funding, conflict of interest, or data availability statements are visible.

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they used generative AI to come up with new ideas for their review.