A single-arm open trial design was used to assess the feasibility and acceptability of a new eHealth intervention aimed at reducing procrastination among college students, which are the primary aims at this stage. This time- and cost-efficient approach allowed for initial insights into engagement and implementation before proceeding to a randomized controlled trial to better examine effectiveness. In this single-arm within-group design, assessments were administered at baseline (t0), post intervention (t2), and follow-up (t2), occurring at 4 weeks and 6 months post baseline, respectively.

The ethical approval was granted by the VU Scientific and Ethical Review Board (reference 2020.088). Participants gave their informed consent, and their privacy and confidentiality were respected through the omission of any identifying details. Participants were not offered compensation for their participation. This study is conducted within the Caring Universities Consortium, which comprises 7 Dutch universities (VU Amsterdam, Leiden University, Maastricht University, Utrecht University, Erasmus University, University of Amsterdam, and Inholland University of Applied Sciences). This consortium is affiliated with the World Health Organization’s World Mental Health International College Student Initiative (WMH-ICS) [51].

This study adheres to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines. A completed checklist is provided in Checklist 1.

There is no standardized method for determining the sample size of an open feasibility study, and previous research has recommended 12 [52] to 35 or more participants [53]. Drawing on these guidelines and comparable studies [54,55], we estimated that 50 participants would be sufficient to address our primary objectives. However, the present intervention was available to the general student population for several years, and we expected that more than 50 participants would sign up. For quality assurance, a power analysis was conducted, and aiming for a conservative 2-tailed calculation with a small effect size (0.2), our statistical analyses (estimating differences between 2 dependent means) would require a total sample size of 199 participants [56]. In this study, we included all participants who had used the intervention and completed the posttest at the time of data analysis to be able to draw firmer conclusions.

Participants were students enrolled in any one of the 7 Dutch universities within the Caring Universities consortium. Recruitment took place in three ways from January 2021 to December 2023. First, through an annual online mental health survey of the WMH-ICS, participants who scored above 28 points on the Irrational Procrastination Scale (IPS) in the survey were invited to participate in the intervention. Second, through social media marketing as well as on-campus marketing via the distribution of posters, flyers, and other promotional materials. Third, college staff members such as student psychologists, study advisors, and lecturers could recommend the intervention to students who might benefit from it.

Participants were eligible if they were (1) aged 16 years or older, (2) enrolled as a student in one of the participating universities, (3) reported no active suicidal ideation, and (4) provided informed consent.

We identified participants with suicidal ideation with a series of questions. Participants who indicated that they had recently had thoughts of killing themselves and were somewhat likely or very likely to act on these thoughts were excluded from the intervention. These participants were provided with an extensive list of possible resources to turn to. This message was also sent to their email address.

The intervention “GetStarted” was created in collaboration with a User Experience designer and followed the principles of an optimized user interface [57]. Based on the principles of CBT, the intervention consisted of 5 mandatory modules and 4 optional modules. The optional modules became available to participants upon completion of the second mandatory module. Participants could choose whether they wanted to do optional modules, which ones, and in which order. An overview of all the modules can be found in Table 1.

The intervention was guided by e-coaches. These e-coaches were trained (research) master’s students in clinical psychology and third-year clinical psychology bachelor students. To be eligible for the position of e-coach, students had to have successfully completed courses on diagnostics and psychological interviewing. Selected candidates completed 6 hours of training before commencing their coaching activities.

When participants enrolled for the intervention, they were given the option to choose an e-coach based on a photo and a short biography of 3 random e-coaches. From then on, the e-coaches supported the participants and provided guidance through asynchronous, text-based feedback delivered via the platform. Feedback was written based on the participants’ answers to the questions posed to them by the intervention. In the feedback, coaches showed empathy and understanding, encouraged the participant to apply the newly learned techniques, and asked reflective questions.

These coaches also attended weekly 1-hour intervision sessions under the supervision of the research team to enhance their coaching skills and maintain quality standards. During each intervision session, several e-coaches would share messages they had sent to their participants and would receive feedback from their peers and the supervisor. The supervisor also checked messages sent by e-coaches on the platform on a sample basis as a form of quality control.

The overview of the assessments can be seen in Table 2.

IBM SPSS version 27 was used for the data analyses.

First, we examined the baseline characteristics of the whole sample, study completers versus noncompleters (ie, participants who did vs did not complete the posttest assessment) and intervention completers versus noncompleters (ie, participants who did vs did not complete at least 4 main modules). Potential differences in baseline characteristics between these groups were examined using χ² tests and independent sample t tests.

For our primary outcomes of client satisfaction (CSQ-8) and usability (SUS-10), we conducted a complete case analysis and analyzed only study completers. We calculated descriptive statistics for the whole sample and examined intervention completers versus noncompleters as an additional sensitivity analysis. For the primary outcome of adherence, we calculated how much of the program was completed on average, as well as the percentage of participants who completed the intervention versus did not complete the intervention.

For our secondary outcomes, which were of an exploratory nature, we conducted a complete case analysis and examined whether there was a statistical difference from baseline to posttest in reported procrastination behavior (IPS), depression (PHQ-9), stress (PSS-10), and quality of life (MHQoL). We conducted 2-tailed paired t tests using a significance level of α=.05 to assess these changes. To interpret the effect size, we calculated the Cohen d, interpreting the benchmarks of 0.2, 0.5, and 0.8 as small, moderate, and large, respectively [73].

A total of 1746 students were assessed for eligibility for the intervention between January 2021 and May 2023. Of these, 657 participants did not receive access to the intervention because they dropped out during the baseline assessment (n=204), dropped out during account creation (n=359), or were excluded due to active suicidal ideation (n=94).

A total of 1089 participants were granted access to the intervention. Of these, 355 (32.6%) did not start the intervention. The reasons for not accessing the intervention were not activating their account from a link they received via email (n=121) or completing zero modules within the intervention and therefore not starting the treatment (n=234). There were no baseline differences between those who started the intervention and those who did not.

A total of 734 participants started the intervention. Our analyses on adherence were conducted based on this sample. Of these 734 participants, 335 completed the posttest (ie, study completers) and were included in the satisfaction and clinical outcome analyses. The details on the participant flow can be found in Figure 1.

The average age of the total sample (n=734) was 23.6 (SD 4.15) years, and most participants were female (n=543, 74%). Most were from the Netherlands (n=446, 60.8%) or another country in Europe (28.6%). Almost half of the participants were currently doing their master’s (n=340, 46.3%). A vast majority of participants (n=615, 83.8%) were currently receiving no medication or psychotherapy. We compared baseline characteristics of intervention completers and noncompleters as well as study completers and noncompleters. Age, gender, nationality, university, education level, marital status, current professional help, and baseline clinical characteristics (procrastination, depression, stress, and quality of life) were compared. We found no significant differences between intervention completers versus noncompleters. When comparing study completers versus noncompleters, the only significant difference we found was the current professional help, with study noncompleters reporting use of medication, psychotherapy, or both more frequently than study completers (P<.001). Full details on the baseline characteristics of the total sample as well as of study completers versus noncompleters and intervention completers versus noncompleters can be found in Multimedia Appendix 1.

Adherence was calculated based on all participants who started the intervention (ie, completed at least 1 module; n=734). In this study, completing 4 main modules meant that the participant completed the intervention, which was done by 269 of 734 (36.65%) participants. This would indicate a dropout rate of 63.35% (n=465); however, participants who dropped out still partially adhered to the intervention. To capture this partial adherence, we calculated that participants completed an average of 2.97 out of 4 main modules, corresponding to 68.95% of the overall intervention.

Many participants also chose to do one or several optional modules. The most popular topics were “productivity” and “motivation boost,” which were completed by 26.7% (n=196) and 24.5% (n=180) of participants, respectively. The optional modules about “task management” and “time management” were completed by 19.5% and 14.9% of participants, respectively.

E-coach satisfaction was assessed among participants who completed the posttest assessment (n=335). The average score on the WAI-I for the total sample was 45.26 (SD 7.72, scores ranging from 15 to 60). Participants who completed the intervention (n=175) reported higher satisfaction (mean 47.39, SD 6.91) than intervention noncompleters (n=160; mean 42.93, SD 0.62). The difference was significant (t₃₃₃=5.52; P<.001; Cohen d=0.6).

This study demonstrates promising outcomes for the feasibility and acceptability of “GetStarted,” an eHealth intervention targeting procrastination specifically designed for college students guided by psychology students. Participants reported high satisfaction with the intervention and found the system usable. Adherence rates were acceptable given the nature of the intervention (self-help), the topic (procrastination), and the target group (students). While most participants did not finish the intervention, on average, they still completed a large part of the intervention. Participants showed a moderate-to-large decrease in procrastination, a small-to-moderate decrease in depression and stress, and a small-to-moderate increase in quality of life between pre- and posttest. However, these results must be interpreted with caution given the lack of a control group in the current study design. Lastly, participants were sufficiently satisfied with the guidance of student e-coaches.

Our findings are in line with previous research on eHealth interventions among college students. Students report being satisfied with interventions for varying mental health problems such as stress, depression, and anxiety [54,74,75]. In this study, satisfaction scores were also found to be acceptable for the total sample and high for intervention completers. Internet interventions for mental health complaints are also generally found to have moderate-to-high usability [54,74,76,77]. Usability scores for the present intervention were high and corresponded with very good–to-excellent user experience [78].

Regarding intervention adherence, a meta-analysis reported dropout rates of up to 50.33% in guided and unguided internet interventions for college students [79]. More recent individual studies of guided internet interventions similar to ours reported dropout rates of 66.4% [74] and 48% [54]. In this study, 63.35% (n=288) of participants who started the intervention did not complete it. While this is on the high side, it is still in line with prior research. Additionally, on average, participants completed nearly 70% of the intervention and therefore were exposed to the larger part of the treatment. It is likely that participants benefited from this exposure to the treatment, even if they subsequently dropped out. We therefore deem our adherence rate on the low side, yet consider it acceptable in order to move forward with the “GetStarted” intervention.

This study also examined any changes to procrastination behavior, depression, stress, and quality of life in an exploratory fashion. While the absence of a control group means we cannot draw any conclusions on treatment effectiveness, we did see an improvement in all these areas. This is also in line with prior research, which has shown that internet interventions can be effective in treating different kinds of mental health problems in students, such as depression, anxiety, stress, and eating disorder symptoms [79] as well as procrastination [30-32]. The implication of our findings is 2-fold. First, this study provides preliminary indications that GetStarted may be a promising eHealth intervention to reduce procrastination behavior in college students, as reflected by a moderate-to-large effect size. Second, our findings tentatively suggest that targeting one specific problem such as procrastination can also potentially reduce other mental health complaints such as depression or stress, even if the intervention was not aimed at these problems specifically. However, it must be noted that spontaneous remission or regression to the norm cannot be ruled out in our study, given our single-arm design. Further research utilizing a control group is needed to draw conclusions on intervention effectiveness.

Lastly, we examined whether guidance by trained psychology students was acceptable. Our findings indicate sufficient satisfaction with the guidance, which is comparable to other guided, internet-based, or face-to-face interventions [74,80,81].

Considering these findings, the new internet intervention “GetStarted” targeting procrastination in college students appears very promising.

A main strength of this study is the comprehensiveness of the assessment. When examining the acceptability and feasibility of new eHealth interventions, studies mainly focus on participant satisfaction and usability. While these were the main outcomes of this study, we also explored changes to procrastination and several other related mental health complaints. Given that procrastination is linked to mental health issues such as depression, social anxiety, stress, and low self-esteem [13,14,16,17], this study provides valuable insights into how treating procrastination may also impact mood, stress, and quality of life. Additionally, we employed posttest as well as follow-up measures. This allowed us to explore both short-term and longer term changes to these mental health complaints.

However, a major limitation of this study is that we cannot draw any conclusions about the effectiveness of the intervention on student mental health due to the lack of a control group. While this feasibility study was not designed to evaluate effectiveness, we did report pre-post change scores on secondary outcomes. While we found positive changes on all the measures, none of these can be directly attributed to the intervention given the lack of a control group. Regression to the mean or spontaneous remission offers alternative explanations for the improvements we found. Moreover, the observed effects may be inflated due to selective study attrition. Although a formal bias analysis falls outside the scope of this study, we acknowledge that the true change may be smaller than reported. Therefore, in future work, we plan to conduct a randomized controlled trial that includes a suitable comparison condition (eg, waitlist control, treatment-as-usual, or active placebo). This will allow us to more rigorously evaluate intervention effectiveness and account for nonspecific factors such as spontaneous remission and placebo effects.

A third limitation of this study regards the sample. During the recruitment process, all students of participating universities were offered access to the intervention. However, not all students are equally likely to be interested in or willing to sign up for the intervention, which means our findings are not directly generalizable to the total student population. For example, the majority of participants in this study were Dutch and female, which limits our understanding of intervention feasibility for other genders and nationalities. As of April 2025, 44.3% of university students in the Netherlands were male and 27.1% were international [82,83], compared to 25.2% (n=185) and 39.2% (n=288), respectively, in our sample. Our sample seems sufficiently diverse in terms of nationality; however, male students are underrepresented. A possible approach for future research is to first try to gain a deeper understanding of why male students in the Netherlands are less likely to make use of the intervention. Prior research has found that male students may perceive fewer benefits of and hold more stigma-related attitudes regarding mental health help-seeking [84]. Additional interviews with this target demographic could yield specific, actionable insights on how to increase the uptake of the “GetStarted” intervention. Additionally, targeted marketing efforts could be made. For example, colleges host many events for specific groups. These events can be a great opportunity to create awareness of the intervention among certain groups. Similarly, online marketing efforts such as social media posts could target certain demographics. A recommendation for future recruitment is to aim for a sample that is between 40% and 45% male participants. A more diverse study sample in further research on the intervention would greatly benefit the generalizability of future findings.

Lastly, the high attrition is a limitation in this study. Less than half of the participants who started the intervention completed the postmeasure, and about a third completed the 6-month follow-up. Though we found no differences in the baseline characteristics of study completers and noncompleters, we cannot rule out any unmeasured confounders. This results in a risk of attrition bias, meaning that “differences between people who leave a study and those that continue can be the reason for any observed effect and not the intervention itself” [85]. Future research should employ strategies to reduce attrition, such as calling participants as a reminder or offering a financial incentive, in order to avoid attrition bias. Both reminder-based approaches and financial incentives have been shown to improve retention in online intervention trials [86,87].

A strength of this study is our exploration of student e-coaches as the source of guidance. To the best of our knowledge, this study was the first to investigate the acceptability of guidance offered by trained psychology students. Prior research has shown that guided internet interventions are more effective than unguided treatment. However, mental health resources such as properly trained professionals are scarce. It is of paramount importance to find scalable and cost-effective alternatives to a licensed psychologist so that people in need can get timely mental health care. Clinical psychology students are required as well as eager to gain hands-on experience during their studies, which makes them a promising potential resource. This study shows that participants of our eHealth intervention are satisfied with the guidance offered by student coaches, and tapping into this more readily available resource means people in need can get more or quicker access to mental health support.

Finally, a last strength of this study is the scalability of our approach, starting with the guidance offered by e-coaches. In this study, e-coaches were involved as a part of their curriculum (ie, a “mini-internship”), meaning they did receive study credits but not financial compensation. This greatly adds to the scalability of our approach. While supervising more e-coaches does require slightly more resources in terms of work hours, this investment is negligible compared to the costs involved with utilizing mental health professionals.

The digital platform used to deliver the intervention was specifically developed with scalability in mind. The platform was created by the Caring Universities Project, which is affiliated with the World Health Organization’s WMH-ICS [51]. We currently offer access to 11 digital interventions, including “GetStarted,” and adding additional interventions incurs minimal additional cost. Participating institutions pay a membership fee in exchange for unlimited access to all interventions. Importantly, there is no cap on the number of students who can enroll. Although there are some ongoing maintenance costs associated with platform upkeep, these remain relatively modest, especially in light of the wide range of interventions and the large number of students that can be reached. Additionally, given the noncommercial nature of Caring Universities, the per-institution fee decreases as additional universities join the initiative, while platform costs remain stable. This model offers substantial potential for large-scale dissemination and implementation.

The internet-based, student-guided intervention “GetStarted” targeting procrastination appears to be acceptable and feasible for college students in the Netherlands. The intervention could potentially also reduce complaints in the areas of procrastination, mood, and stress. Future research should determine possible intervention effectiveness using a control group, as well as explore strategies to address high study attrition.