Table 6.
Factors associated with treatment engagement and outcome.
| Study | Intervention and comparator | Adverse events | Treatment nonuptake | Treatment dropout/attrition | Treatment engagement, satisfaction, and therapeutic alliance | Is there an association between treatment adherence and treatment outcome? | Factors associated with treatment engagement and treatment outcome | Other findings relevant to treatment engagement and treatment outcome |
|---|---|---|---|---|---|---|---|---|
| Burns et al. [56] | C/BT vs. routine clinical care | Not reported | Unclear | Intervention 2 of 18 (11.11%) Control 5 of 18 (27.78%) Reasons given for withdrawal from CBT included being too busy and disliking CBT. |
Good engagement, completion rate of 72% | Not reported | (−) Baseline depression score and improvements in depression scores. (+) Improvement in depressive symptoms and easier infant temperament and shorter nocturnal sleep duration, (n = 25). |
Not reported |
| Carter et al. [58] | C/BT vs. routine clinical care | Not reported | Intervention 66.67% Control 0%. Reasons for nonuptake of C/BT treatment were reported as feeling they did not need treatment. |
Intervention 2 of 3 (66.67%) Control 0 of 3 (0%) |
Good engagement for the one individual who received, 100% completion rate. | Not reported | (+) Women who were more symptomatic on the EPDS were more likely to agree to be contacted to be considered for participation in the study (only six individuals were randomized). | Not reported |
| Cho et al. [59] | C/BT vs. psychoeducation | Not reported | Intervention 0% Control 0% |
Intervention 3 of 15 (20%) Control 2 of 12 (16.67%), with reasons, including premature delivery and no longer feeling depressed |
Good engagement, 80% completion rate. | Not reported | Not reported | The intervention included the review and modification of dysfunctional cognition in the marital relationship of the parents. This improved through treatment. |
| Dimidjian et al. [48] | C/BT BA plus TAU care vs. TAU | Not reported | Intervention 6.98% Control unclear |
Intervention 8 of 86 (9.30%) Control 5 of 77 (6.49%) |
Good engagement, two-thirds of participants completing five or more sessions (M 6.43; SD = 3.64). High degree of satisfaction on the CSQ-8 (M 27.76; SD = 3.83). |
Not reported | Not reported | Early change in self-reported putative targets was found to predict later improvement in symptoms of depression (NB - the causal relationship cannot be inferred through the methodology, which was underpowered). |
| Forsel et al. [60] | i-C/BT GSH vs. TAU with option to receive i-C/BT at a later date | Two patients reported one minor adverse event, which was that they felt stressed about not keeping up with the treatment program. Deterioration (considered as 4 points or more increase on the primary outcome MADRS-S) was observed in three participants in TAU and one participant in i-CBT but the difference was not significant. | Intervention 4% Control unclear |
Intervention 1 of 22 (4.54%) Control 2 of 20 (10%) |
Good engagement. They completed on average 5.3 modules (SD = 2.5) but were given on average of 7 (SD = 2.1) out of a possible 10 modules. TCS mean score was 33.8 (SD = 9.1) 2 weeks into treatment. CCSQ mean score was 23.8 (SD = 3.1), indicating good satisfaction. |
Not reported | Not reported | All but one patient believed that the pregnancy-adaptation of the treatment was important and helpful. |
| Hayden et al. [61] | C/BT vs. supportive counseling | Not reported | Unclear | Unclear | Average therapy duration was 7.7 weeks (SD = 2.8), of a total of 10 weeks, and duration was not different between therapy groups. | Not reported | (−) Treatment outcome and participant demographics (age, race). (−) Attrition and demographics. (−) Attrition and baseline depression. |
Not reported |
| Milgrom et al. [62] | C/BT vs. routine clinical care | Not reported | Intervention 11.11% Control unclear Reasons for nonuptake included one individual being hospitalized shortly after joining the study and suffered a reproductive loss at 25 weeks gestation, and one cited other health concerns as a reason. |
Intervention 4 of 27 (14.81%) Control 6 of 27 (22.22%) |
Of a possible eight sessions, women in the intervention attended an average of 6.30 sessions (SD = 2.91). The helpfulness of the intervention was rated 8.6/10 (SD = 1.4), and women's satisfaction with their treatment was rated 9/10 on average (SD = 0.9). All respondents to these questions (n = 19) reported that the intervention had been sufficient to address the problems they had been facing. |
(+) There was a strong dose–response relationship between number of C/BT sessions and amelioration of depression. | Not reported | Not reported |
| Edge [39] | i-C/BT GSH vs. TAU | Methodology for measurement described, and their absence was also reported. | Intervention 0% Control 0% |
Intervention 2 of 21 (9.52%) Control 1 of 22 (4.55%) |
Good engagement with 86% (18 of 21 participants) completing all six sessions. Women visited the program on a mean of 20.5 occasions (SD = 10.6) and the mean number of sessions attended was 5.7 (SD = 0.7). Mean satisfaction ratings were in the moderately satisfied range (mean 3.1, SD = 0.60, range 2–4) on a 4-point scale. Similarly, mean ratings of the helpfulness of telephone coach calls were in the moderately helpful range (mean 3.2, SD = 0.89, range 1–4) on a 4-point scale. |
Not reported | Not reported | Partners were signposted to resources, in recognition of their role, with 76% accessing this resource, however no improvement was found in women's relationships with their partners over the study period. |
| Milgrom et al. [64] | i-C/BT GSH vs. TAU | Methodology for measurement described, but no adverse events reported. | Intervention 3% Control unclear |
Intervention 7 of 39 (17.95%) Control 3 of 39 (7.90%) |
72% of participants (28 of 39) viewed all six sessions. Of the women who visited at least once, the mean number of sessions viewed was 5.6 (SD = 1.7). The helpfulness of the coach calls and treatment were rated on a scale from 0 to 3, with ratings of 2.56 (SD = 0.67), and 2.31 (SD = 0.89), respectively. |
Not reported | (−) No association was found between attrition and the following baseline variables: depression; anxiety; perceived stress; marital functioning; CBT skills (Automatic Thoughts Questionnaire, Behavioral Activation for Depression Scale); maternal self-efficacy. |
Participants and their partners were pointed towards a library of resources with content on PND and parenting. Participants rated the helpfulness of library articles and videos on a scale of 0 to 3, with a mean of 1.96 (SD = 1.04), and 1.87 (SD = 1.07), respectively. Partners were pointed to a partner support website and 39% (15 of 39 partners) accessed this, with participants rating the helpfulness of the partner website on a scale of 0 to 3, with a mean of 1.19 (SD = 1.02). |
| Ngai et al. [17] | Telephone-based C/BT vs. TAU | Not reported | Intervention 5.58% Control 0% |
Intervention 10 of 197 (5.08%) Control 2 of 200 (1%) |
86.8% completed all five sessions. | Not reported | Not reported | Parenting stress, quality of life, parenting self-efficacy and satisfaction were measured in secondary analysis. Improvements were made in these areas for the group receiving C/BT. |
| O'Mahen et al. [68] | C/BT vs. TAU | Not reported | Intervention 17% Control unclear |
Intervention 8 of 30 (26.67%) Control 4 of 25 (16%). Reasons for dropout included not liking the session-by-session questionnaires, wanting more practical advice from the therapist, difficulties prioritizing self over family, and childcare difficulties. |
Women received an average of 2.30 (SD = 2.16) sessions during pregnancy, and 5.35 (SD = 4.07) postpartum, of a total of 12 sessions. Women reported high levels of content applicability and treatment satisfaction in session-by-session assessments which used open-ended questions. |
(+) Both treatment adherence and treatment satisfaction were associated with a reduction in depression scores. | (+) Applicability of the treatment material was associated with treatment satisfaction. (−) Applicability of the treatment material was not associated with depression at posttreatment, after controlling for depression at baseline. (+) Women more functionally impaired and women with higher baseline depression scores were less likely to engage with treatment. (−) Other participant demographic factors were not associated with treatment engagement (factors not clearly stated). (−) Participant psychological barriers to help seeking, such as avoidance (assessed via a 25-item measure with a 5-point Likert scale), were not associated with treatment engagement. |
Barriers to adherence included struggling with care demands of a new baby, child illnesses and pregnancy-related pain, housing concerns, and lack of a private, safe home in which to meet. |
| O'Mahen et al. [69] | i-C/BT GSH vs. TAU | Not reported | Unclear | Intervention 281 of 462 (61.8%) Control 286 of 448 (63.84%) |
Engagement was low, overall. There was good initial adherence reported in terms of session views, but this reduced significantly over time, with the greatest drop happening between sessions 1 and 2. | Not reported | (+) Women with a higher socioeconomic status were more likely to complete the baseline questionnaire. Qualitative information suggesting a link between hopelessness associated with depression and likelihood of adhering to treatment given that the hopelessness may permeate a belief around getting better. |
There was a low response (22%) to a link to acceptability questions, but responses revealed acceptability and a role for internet-based interventions for women struggling with stigma and practical barriers to treatment. The qualitative study revealed that the anonymity of internet-based approaches was helpful in terms of stigma, but that stigma might still stop mums from engaging in homework practical activities. It also found that the perceived lack of social support as a mum is overwhelming, and a barrier to seeking help. The qualitative study (2015) highlighted themes, including the importance of guidance, and of personalized and relevant content, with examples that mums can relate to. The importance of accessibility was also raised, whilst noting that some would prefer a more structured routine particularly if they are lacking in motivation. Some noted a preference for face-to-face. |
| O'Mahen et al. [71] | i-C/BT GSH vs. TAU | Not reported | Intervention 2.44% Control unclear |
Intervention 3 of 41 (7.32%) Control 8 of 42 (19.05%) |
Women viewed a mean of 6.74 sessions (SD = 4.53), and completed 5.36 sessions (SD = 4.62), of a total of 12 sessions. | Not reported | (−) Telephone session adherence was not associated with the participant demographics of income, work status, relationship status, academic qualifications, number of children, or ethnicity. (+) Women with lower perceived support completed fewer program modules. (+) Women who were working or studying for a degree completed fewer program modules. (+) Women with poorer work and social adjustment baseline functioning and who were of a lower socioeconomic status opened fewer sessions. |
The researchers note that women with complex life circumstances required more time to work through materials and supporters helped in terms of pacing the treatment content. |
| Pugh et al. [72] | i-C/BT GSH vs. waitlist | Not reported | Intervention 4% Control 0% |
Intervention 4 of 25 (16.67%) Control 3 of 25 (12%) |
Participants completed on average 5.92 of the seven modules (60% of the participants completed all seven modules). Over 80% of the participants reported that they liked the overall program. Participants also reported a high level of therapeutic alliance, giving a rating of 86.42% (TSQ) |
Not reported | Not reported | Not reported |
| Rondung et al. [73] | i-C/BT GSH vs. TAU | Not reported | Intervention 19% Control unclear. Researchers note quite a few women declined participation because they did not accept randomization to either intervention, that is, they preferred standard care. |
Intervention 43 of 127 (33.86%) Control 26 of 131 (19.85%) |
Very low adherence in the guided i-C/BT group, with fewer than half of participants going on to the second module and less than one-third advancing to the third. | Not reported | (−) Participants defined as lost to follow-up were no different from the others who were not lost to follow-up, with regard to any preintervention characteristic or the level of fear of birth at screening. | The qualitative study [74]), reported view of i-C/BT being meaningful, helpful, flexible, private, but also not pinpointing fears, demanding, and resulting in feeling alone. The pregnancy itself brought an increased load affecting both family-life, work and the women's health which contributed to a low motivation to complete the i-C/BT, though some were motivated towards using i-C/BT. The range of views reported in the qualitative work suggest the importance of guidance and tailoring guidance. |
| Shaw et al. [76] | C/BT vs. Information/TAU | Reported as none | Intervention 0% Control 4.65% |
Intervention 5 of 62 (8.06)% Control 2 of 43 (4.65%) |
90% of mothers completed all six sessions and were satisfied with the treatment (satisfaction measure not noted). |
Not reported | (+) Mothers with higher ratings of baseline NICU stress benefited more from the intervention compared with mothers who had lower ratings. (+) Mother's education and household income were found to be nonspecific predictors of the outcome (that is, they predicted outcome regardless of which intervention they received). Less educated mothers and mothers with lower household income showed a greater decline in trauma symptoms irrespective of whether they were in the intervention or comparison group. (−) The following variables were not found to be associated with treatment outcome: infant illness severity index, length of NICU stay, Traumatic Events Questionnaire, current major depressive episode, maternal age, white/non-white race, Hispanic ethnicity, and US born. |
The researchers note that mothers in general reported greater satisfaction with the shorter than typical treatment (around 6 instead of 12 sessions), due to demands associated with having their infant in NICU. |
| Suchan, et al. [78] | i-C/BT GSH vs. TAU | Methodology for measurement described, and their absence was also reported. | Intervention 6.67% Control 3.03% |
Intervention 4 of 30 (13.33%) Control 2 of 33 (6.06%) |
75% (21/28) of the participants completed at least four of the five lessons. Pretreatment mean credibility factor score of the CEQ was 21.22 (SD = 3.38), and the mean score on the expectancy factor was 17.07 (SD = 3.77), with no differences between the treatment groups. Clients in the intervention treatment group demonstrated a significant increase in treatment credibility scores after treatment, mean of 23.58 (SD = 3.02). High levels of alliance with therapists according to the WAI. |
Not reported | Not reported | Not reported |
| Trevillion et al. [79] | C/BT GSH plus TAU vs. TAU | Nine were reported but were not related to mental health or involvement in the trial. | Intervention 11.54% Control unclear |
Intervention 2 of 26 (7.69%) Control 1 of 27 (3.70%). Reasons for withdrawal from treatment included inconvenience of the intervention alongside existing commitments, and symptoms resolved. |
18 participants (69%) attended at least the minimum number of sessions (≥4 sessions). The overall median number of sessions attended was 6.5. Therapists demonstrated competence with regards to interpersonal effectiveness when fidelity was checked. |
Not reported | Not reported | Not reported |
| Wozney et al. [80] | C/BT GSH vs. TAU | Not reported | Unclear | Intervention 5 of 32 (16.63%) Control 2 of 30 (6.67%) |
The mean number of sessions across all intervention participants was 9.13 sessions (SD = 4.59), of a total of 12. A 10-item satisfaction questionnaire created by the research team was used and high satisfaction was found, including the helpfulness of coaches. |
(+) A dose–response relationship between treatment engagement and outcome. Participants who engaged with a higher number of the intervention sessions were significantly more likely to no longer meet diagnostic criteria for depression. For every additional session of the intervention received, individuals had a 1.4 times greater chance of showing improvement. | (−) Other potential moderators, including cointervention, psychotropic medication, treatment duration, and booster sessions were not associated with treatment outcome. | Not reported |
Note: (+) indicates an association and (–) indicates no association.
Abbreviations: BA, behavioral activation; C/BT, cognitive and/or behavioral therapy; CEQ, credibility/expectancy questionnaire; CSQ-8, clients satisfaction questionnaire-8; EPDS, Edinburgh postnatal depression scale; i-C/BT, internet-based CBT; MADRS-S, Montgomery and Asberg depression rating scale; NICU, neonatal intensive care unit; TAU, treatment as usual; TCS, treatment credibility scale; TSQ, treatment satisfaction questionnaire; WAI, working alliance inventory.