What this article is about
In brief
This is the 2012 study protocol for Mobile. Net, a Finnish trial testing patient-tailored SMS reminders to support medication adherence and follow-up after psychiatric hospital discharge; it describes the planned design, not results.
For nursing students
Study summary
This article is a study protocol, not a results paper. It describes how researchers in Finland planned to test whether text messages sent to a mobile phone could help people with psychosis take their medication and keep follow-up appointments after leaving hospital. The trial is called Mobile.Net, and it was designed by Maritta Välimäki, Heli Hätönen, and Clive E Adams and published in JMIR Research Protocols in 2012.
Schizophrenia and related psychotic illnesses are costly, long-term conditions, and the authors note that simple, low-cost tools to support people after discharge were missing. Text messaging was chosen because mobile phones are widely available and a message can be personalized without requiring extra staff time for every contact.
The plan was a pragmatic, multicenter randomized controlled trial across 24 sites and 45 psychiatric hospital wards in Finland. Whole wards were randomized to one of two arms, and participants within each ward were then allocated individually using a computer-generated permuted block design with sealed, sequentially numbered envelopes handled by research nurses. Eligible participants were adults aged 18 to 65 being discharged from a psychiatric hospital on antipsychotic medication, who owned a mobile phone, could use Finnish, and could give informed consent. People needing only planned non-acute care or those in forensic psychiatric services were excluded. The protocol set a recruitment target of roughly 1,511 participants per arm, calculated to have 80% power to detect a 5-percentage-point reduction in hospital readmission.
In the intervention arm, participants received semiautomatic SMS messages after discharge. What made this different from many earlier text-message studies is that patients themselves chose the wording, topic, timing, and frequency of the messages, covering things like medication reminders, appointment reminders, and general daily care. Messages continued for up to 12 months, or stopped earlier if a participant asked to withdraw. Everyone in both arms kept receiving whatever usual psychiatric and nursing care their team normally provided; the SMS messages were added on top of standard care for the intervention group only.
The primary outcomes planned were service use and "healthy days" over 12 months, drawn from Finland's national health registers: whether someone was readmitted to hospital, how long it took before a first readmission, and total days spent in hospital during the year. Secondary outcomes included further service use, use of coercive measures (such as restraint or seclusion), medication-related data, adverse events, satisfaction with care and with the trial itself, social functioning, and cost. The researchers planned to analyze the results by intention-to-treat, meaning participants would be analyzed according to the group they were randomized to, regardless of how much they actually used the messages. The trial was registered publicly (ISRCTN27704027) and received ethics approval from the Hospital District of Southwest Finland in December 2010.
Because this is a protocol article, it reports no outcome results — it only describes what the researchers intended to measure and how. For nursing students, this is a useful example of how a rigorous mental-health intervention trial is designed before any data are collected: defining eligibility clearly, choosing a randomization method that limits bias, deciding outcomes in advance, and planning the statistical approach ahead of time. It also shows an interesting design choice worth discussing in class: letting patients control the content of their own reminder messages, which is meant to increase acceptability and reduce the feeling of being "nagged" by a rigid, one-size-fits-all system. Students appraising this article should notice what is, and is not, claimed: a protocol tells you the plan, not whether the plan worked.
Original publication
Source abstract and study details
Read the source abstract
BackgroundSchizophrenia is a high-cost, chronic, serious mental illness. There is a clear need to improve treatments and expand access to care for persons with schizophrenia, but simple, tailored interventions are missing. ObjectiveTo evaluate the impact of tailored mobile telephone text messages to encourage adherence to medication and to follow up with people with psychosis at 12 months. MethodsMobile.Net is a pragmatic randomized trial with inpatient psychiatric wards allocated to two parallel arms. The trial will include 24 sites and 45 psychiatric hospital wards providing inpatient care in Finland. The participants will be adult patients aged 18–65 years, of either sex, with antipsychotic medication (Anatomical Therapeutic Chemical classification 2011) on discharge from a psychiatric hospital, who have a mobile phone, are able to use the Finnish language, and are able to give written informed consent to participate in the study. The intervention group will receive semiautomatic system (short message service [SMS]) messages after they have been discharged from the psychiatric hospital. Patients will choose the form, content, timing, and frequency of the SMS messages related to their medication, keeping appointments, and other daily care. SMS messages will continue to the end of the study period (12 months) or until participants no longer want to receive the messages. Patients will be encouraged to contact researchers if they feel that they need to adjust the message in any way. At all times, both groups will receive usual care at the discretion of their team (psychiatry and nursing). The primary outcomes are service use and healthy days by 12 months based on routine data (admission to a psychiatric hospital, time to next hospitalization, time in hospital during this year, and healthy days). The secondary outcomes are service use, coercive measures, medication, adverse events, satisfaction with care, the intervention, and the trial, social functioning, and economic factors. Data will be collected 12 months after baseline. The outcomes are based on the national health registers and patients’ subjective evaluations. The primary analysis will be by intention-to-treat. Trial RegistrationInternational Standard Randomised Controlled Trial Number (ISRCTN): 27704027; http://www.controlled-trials.com/ISRCTN27704027 (Archived by WebCite at http://www.webcitation.org/69FkM4vcq)
Reviewed findings
Main findings
- The protocol describes Mobile.Net as a pragmatic, multicenter randomized controlled trial across 24 sites and 45 psychiatric hospital wards in Finland, testing tailored SMS text messages against usual care.
- Eligible participants were adults 18-65 with antipsychotic medication at discharge from a psychiatric hospital, owning a mobile phone, able to use Finnish, and able to give informed consent.
- The intervention let participants choose the content, timing, and frequency of their own SMS reminders covering medication, appointments, and daily care, continuing up to 12 months or until withdrawal.
- Randomization used a computer-generated permuted block design with sealed, sequentially numbered envelopes, with wards allocated to arms and participants individually randomized within wards.
- Primary outcomes were hospital service use and 'healthy days' at 12 months from national health registers, with secondary outcomes covering coercive measures, medication, adverse events, satisfaction, social functioning, and cost, analyzed by intention-to-treat.
Using the findings
Practice considerations
- Nurses supporting discharge planning can consider that patient-controlled reminder content (rather than fixed, generic messages) was the design feature this team believed would improve acceptability of a text-messaging support tool.
- Because this is a protocol, not a results paper, nurses should not assume text messaging improves adherence or reduces readmission based on this article alone; effectiveness has to be judged from the completed trial's outcome data.
- The trial's reliance on national health-register data for hospitalization outcomes illustrates how administrative data can supplement self-report in evaluating adherence-support interventions in real practice settings.
- The inclusion of coercive measures and adverse events as secondary outcomes reflects appropriate safety monitoring practice when introducing any new communication-based intervention into psychiatric aftercare.
- The exclusion of forensic psychiatric patients and those needing only planned non-acute care is a reminder that adherence-support interventions studied in one population may need separate evaluation before extending to other psychiatric populations.
For educators
Teaching and appraisal notes
This is the trial protocol for Mobile.Net, a pragmatic, multicenter, two-armed randomized controlled trial (Välimäki, Hätönen, and Adams, JMIR Research Protocols, 2012; ISRCTN27704027) evaluating tailored SMS text messaging to support medication adherence and post-discharge follow-up for adults with psychosis in Finland. It is well suited for teaching study design and critical appraisal of protocol papers, since it separates planned methodology from eventual results.
Design: the trial randomized whole psychiatric wards (24 sites, 45 wards) to intervention or control, then individually randomized participants within each ward using a computer-generated permuted block design with sequentially numbered, sealed envelopes managed by research nurses — a cluster-then-individual approach intended to limit contamination between arms while preserving individual-level allocation concealment. The protocol reports a target sample of approximately 1,511 participants per arm (roughly 3,022 total), powered at 80% to detect a 5-percentage-point absolute reduction in hospital readmission (relative risk 0.92) at the 5% significance level — useful for a discussion of a priori power calculations and what a 'pragmatic' trial design trades off versus efficacy trials.
Population: adults 18-65, discharged from psychiatric hospital wards on antipsychotic medication, owning a mobile phone, Finnish-speaking, and capable of informed consent; those in forensic psychiatric services or receiving only planned non-acute care were excluded. This is a reasonably representative discharge population for Finnish adult mental health services, though the Finnish-language and mobile-phone-ownership requirements narrow generalizability.
Intervention: a patient-led, semiautomatic SMS system in which participants chose message wording, topic (medication, appointments, daily care), timing, and frequency, continuing up to 12 months or until withdrawal. Both arms continued to receive usual psychiatric and nursing care; blinding of participants and clinical staff was not attempted (open, pragmatic design), but the protocol specifies that outcome assessors and analysts were kept masked to allocation — worth highlighting as a partial-blinding strategy appropriate to a behavioral/service intervention.
Outcomes: primary outcomes were hospital service use and 'healthy days' at 12 months (admission, time to next hospitalization, total inpatient days) drawn from Finnish national health registers; secondary outcomes spanned coercive measures, medication data, adverse events, satisfaction, social functioning, and economic factors. Analysis was planned as intention-to-treat. A notable protocol-stage limitation the authors flagged themselves is a roughly one-year lag in national register data, which precluded any planned interim analysis.
For faculty use: this protocol pairs well with the eventual results paper (Välimäki et al., J Med Internet Res 2017;19(7):e245; PMC5529737), which reported 1,139 participants randomized and found the intervention did not significantly reduce readmission (43.0% vs 38.8%, RR 1.11, p=0.28) or improve most secondary outcomes, though SMS-group participants who were readmitted had somewhat better global functioning, and overall costs were higher in the SMS arm. Comparing protocol to outcome paper is a strong exercise for teaching trial registration, outcome-switching checks, and the gap between a promising, well-designed pragmatic trial and a null clinical result — and for discussing why 'ineffectual' findings are still valuable evidence for practice and policy in digital mental health.
Critical appraisal
Limitations
- As a protocol paper, no outcome results are reported here; conclusions about whether the intervention worked cannot be drawn from this article.
- The authors themselves flagged a roughly one-year lag in national register data, which prevented the planned interim analysis during the trial.
- Eligibility required Finnish-language ability and mobile phone ownership, which may limit how well the planned sample generalizes to non-Finnish-speaking or lower-resource populations.
Classroom use
Discussion Questions
- Why is it important to distinguish a trial protocol paper, like this one, from a results paper when critically appraising evidence?
- What are the advantages and risks of letting patients choose the content, timing, and frequency of their own SMS reminders rather than using a fixed message schedule?
- How does randomizing whole psychiatric wards, then individuals within wards, help reduce contamination between the intervention and control groups, and what risks remain?
- Why might outcome assessors and analysts be masked to allocation even when participants and clinical staff are not, in a pragmatic trial like this one?
- What ethical considerations arise when recruiting adults with psychosis, at the point of hospital discharge, into a randomized trial?
- How might reliance on national health-register data for outcomes like readmission and hospital days differ from relying on self-report, and what are the tradeoffs?
Source-based questions
Frequently asked questions
What is the Mobile.Net trial about?
Mobile.Net is a Finnish multicenter randomized controlled trial testing whether tailored SMS text messages, chosen in content and timing by the patients themselves, help people with psychosis stick to medication and keep follow-up appointments after hospital discharge.
Is this article a study protocol or a results paper?
It is a study protocol, published in 2012, describing the planned design and methods before the trial's outcomes were known; it does not report whether the intervention worked.
Where was this trial registered, and what was its registration number?
It was registered with the International Standard Randomised Controlled Trial Number registry as ISRCTN27704027.
Did the completed Mobile.Net trial ultimately show that SMS reminders reduced hospital readmission?
According to the later published results (a separate 2017 paper from the same research group, not this protocol article), the SMS intervention did not significantly reduce readmission and was described by the authors as ineffectual for the main clinical outcomes, though readmitted SMS-group participants had somewhat better global functioning.