Secondary prevention by structured semi-interactive stroke prevention package in INDIA (SPRINT INDIA): Findings from the process evaluation of a randomized controlled trial

Abstract

Introduction:

Secondary Prevention by Structured Semi-Interactive Stroke Prevention Package in INDIA Trial delivered secondary stroke awareness intervention to sub-acute stroke patients in form of workbook, videos and SMS across 31 centres in 12 languages. Trial was stopped for futility due to fewer vascular outcomes than anticipated. Trial results indicated that trial intervention, did not lead to reduction in vascular events. We carried out process evaluation, to evaluate trial implementation and participant’s perspectives, to comprehend the trial’s futile outcomes.

Materials and Methods:

Using mixed methods approach, qualitative interviews and quantitative data from case report forms, workbooks and questionnaires were analysed to measure intervention fidelity and contamination. Using purposive sampling, 115 interviews of patient-caregiver dyads and health professionals at 11 centres and 2 focus group discussions were held.

Results and discussion:

Iterative thematic analysis of qualitative data was done with RE-AIM and realist models. There was good fidelity to intervention and adherence to protocol; however, there was dilution of inclusion criteria by randomly enrolling uneducated and caregiver-dependent patients. Centre coordinators provided counselling to both arms, not specified by protocol, causing bias. Coordinators found it difficult to keep patients motivated to view intervention which was corroborated by fidelity questionnaire showing decreased viewing of intervention for a year. Cardiovascular protection improved in routine care by virtue of participating in trial. No contamination of intervention was reported.

Conclusion:

The intervention was acceptable by patients and caregivers, which could be made a community-based programme. Reasons identified for decreased viewing were repetitive content and non-availability of personal cellular device.

Introduction

Stroke is second largest cause of deaths worldwide. ¹ In India, stroke is fourth leading cause of death and fifth leading cause of disability with an incidence between 105 and 152/100,000 people per year.^2,3 According to population and hospital-based registries in India, burden of stroke is significantly influenced by high recurrent stroke rates of 15%–21%.^4–6 The most common risk factors of stroke, identified worldwide, are hypertension, diabetes mellitus, smoking, insufficient exercise, cardiac issues, dyslipidaemia, obesity and alcohol intake. ⁷ In low and middle-income economies like India, secondary stroke prevention is challenging due to various reasons including lack of medications, poor drug compliance, low awareness and out-of-pocket costs for treatment. ⁸

To bridge these gaps in stroke care, various approaches for providing quality care and utilizing modern, innovative and culturally appropriate interventions have been debated upon, with a focus on context-specific undertakings. ⁹ mHealth is defined as ‘medical and public health practise supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistants and other wireless devices’ by the Global Observatory for eHealth of the World Health Organisation. ¹⁰ A scoping review conducted on process evaluations (PEs) of complex stroke rehabilitation and secondary prevention trials revealed that half of these interventions were done with mHealth component establishing the future direction of non-communicable diseases prevention. ¹¹ Secondary Prevention by Structured Semi-Interactive Stroke Prevention Package in India (SPRINT INDIA) is the first trial to study the role of a mHealth intervention in stroke secondary prevention on such a large scale in India and perhaps globally. ¹²

SPRINT INDIA, is a randomized controlled clinical trial for sub-acute stroke patients across 31 centres in India under the Indian Stroke Clinical Trial Network (INSTRuCT) from 2018 to 2021 which randomized 4298 patients. ¹³ The intervention arm received SMS messages, videos and a printed workbook on stroke prevention to read and view. ¹⁴ To deliver the SMSs and videos to the intervention arm patients, in bulk every day, the trial leveraged the services of a messaging platform (https://gupshup.io). There was a fixed number embedded in the messages, to which the patient or the caregiver, dialled, after receipt of the message. This call was recorded in the platform as an acknowledgement of the receipt of the message.

The workbook included interactive stroke prevention worksheet which incorporated activities related to that chapter either as questions, fun dice games, treasure hunt on map etc. which were to be completed by patients or caretakers with family and friends. A call was placed to every patient, once a month, to inquire about specified outcome events by centre coordinators. ¹⁴ Follow-ups were conducted at 6 months and 1 year post randomization. ¹⁴

An interim futility analysis by data and safety monitoring board stopped the trial for ineffectiveness due to lower conditional power and fewer primary outcomes of vascular events than anticipated. Trial results indicated no significant difference between intervention and control group in the primary composite outcome. ¹³ Important questions remain about future of stroke prevention package using SMS messages, videos and a workbook for secondary stroke prevention method and the uncertainty as to what transpired on the ground: was the trials result due to failure of the intervention or of trial implementation or both.

Aims

To determine how, for whom and why this complicated intervention had (or did not have) an impact. Our process evaluation (PE) was aimed to be carried out concurrently with the SPRINT INDIA Randomised Clinical Trial (RCT).

Our objectives were (1) to determine whether the SPRINT INDIA trial was implemented as planned; (2) to comprehend how healthcare professionals, patients and carers perceived the impact of care they received or provided; (3) to assess reach of intervention amongst population; (4) to determine how, the trial intervention integrated into the treatment plan, designed to encourage behaviour change and which intervention was more beneficial (5) to determine whether findings are likely to be broadly applicable, sustainable and adoptable.

Methods

PE methodology details are presented in the published protocol. ¹⁵

SPRINT INDIA PE framework

Two theoretical frameworks presented with the Medical Research Council’s guidance addressing continuation of evidence generation-to-practice translation served as foundation for design of our mixed methods evaluation. ¹⁶ RE-AIM model is established on elements – reach (R), effectiveness (E), adoption (A), implementation (I) and maintenance (M), to assess long-term efficacy for both the organizational sector and target subjects. ¹⁷ Realist evaluation provides a rationale for causal outcome patterns in the context of a complex intervention like SPRINT INDIA using Context-Mechanisms-Outcomes. ¹⁸ PE framework and design are presented in Figure 1. ¹⁶

Figure 1.

SPRINT INDIA process evaluation framework.

Data sources

We gathered quantitative process data from 31 centres on delivery of interventions and intervention uptake. These were obtained from screening logs, case report forms, fidelity questionnaires, contamination questionnaires, workbooks and acknowledgement of SMSs and videos receipt data.

The qualitative data were gathered from interviews at 11 centres and focus group discussions (FGDs). The primary reason for selecting these centres is because the intervention was translated into 11 Indian languages and these centres each represent the stakeholder’s languages respectively in the trial. ¹⁹ Hospitals were further categorized for private and government facilities, north and south of India and older and more recent centres that joined the study throughout time. FGDs were conducted with INSTRuCT Network team and Indian Council of Medical Research officials along with health professionals. Patient and caregiver dyads were interviewed following a maximum variance purposive sampling strategy for age, gender, treatment group and region and modified Rankin scale. In-depth interviews of health professionals including centre investigators and coordinators and neutral health professional like a stroke nurse/physiotherapist etc., were also done at these centres. Interviews were conducted using semi-structured interview guides.

Conducting interviews

Interviews were planned to be held in parallel with data collection process of the SPRINT INDIA trial from August 2021 to January 2022. By the time, the trial was stopped for futility in November 2021, interviews at six centres were completed. Interviews at the remaining five centres were conducted, thereafter, till March 2022. Characteristics of sampled centres and interview details are presented in Table 1. Due to COVID-19 pandemic, Zoom interviews were conducted at few centres using telephonic informed consent documents with 17 participants.

Table 1.

SPRINT India process evaluation centres details.

	Centre	Government/Private	North/South	Previous/Newer,*	Language	Mode of conduct
1	Christian Medical College and Hospital, Ludhiana	Private	North	Previous	Punjabi	Onsite – one to one interaction
2	Vadodara Institute of Neurosciences, Vadodara	Private	North	Previous	Gujarati	Onsite – one to one interaction
3	Lalitha Super Speciality Hospital, Guntur	Private	South	Previous	Telugu	Onsite – one to one interaction
4	Kasturba Medical College and Hospital, Manipal	Private	South	Previous	Kannada	Onsite – one to one interaction
5	Bangur Institute of Neurosciences, Kolkata	Government	North	Previous	Bengali	Onsite – one to one interaction
6	All India Institute of Medical Sciences, Bhubaneswar	Government	North	Newer	Oriya	Onsite – one to one interaction
7	Jawaharlal Institute of Postgraduate Medical Education and Research, Puducherry	Government	South	Previous	Tamil	Onsite – one to one interaction
8	Grant Government Medical College and Sir J.J. Group of Hospitals College, Mumbai	Government	North	Previous	Marathi	Onsite – one to one and via Zoom calls interaction
9	Fortis Escorts Hospital, Jaipur	Private	North	Newer	Hindi	Onsite via – telephonic calls
10	Guwahati Neurological Research Centre, Dispur	Private	North	Previous	Assamese	Interaction via Zoom calls
11	Sree Chitra Tirunal Institute of Medical Sciences and Technology, Thiruvananthapuram	Government	South	Previous	Malayalam	Onsite – one to one interaction

^*Previous centres: These centres were part of the SPRINT INDIA from the initial (2018) period. Newer centres: These centres joined the SPRINT INDIA from 2020 onwards.

Interviews were performed by the network’s team, comprising of SJV, DA and AD, who had no previous contact with patients and caregivers, in the stakeholder’s language at each of the 11 centres. Interpreters were used at eight centres where the CCC team members did not speak the regional language.

In-depth semi-structured interviews of 115 stakeholders were executed of 32 health professionals (10 investigators, 11 coordinators and 11 neutral health professionals), 42 patients and 41 caregivers (16 control arm and 25 intervention arm dyads). Interviews were audio recorded and lasted an average of 25 min and transcribed verbatim. Fundamentals of qualitative research to increase rigour and reliability were followed in all aspects as described in the recent publication. ²⁰

Analysis

The interviews in all languages were translated and transcribed by professional translators into English, which were uploaded on NVivo software (release 1.6.2). Qualitative analysis was undertaken by the researchers (SJV and RJI). SJV is pursuing PhD programme in the PE of SPRINT INDIA Trial and has undertaken training to perform thematic analysis. RJI has previous experience of performing thematic analysis for another study. Text was given multiple reads and emergent codes were identified independently to define the interviews of each group of stakeholders. Researchers came together at this point and the various codes were organized to advanced codes to develop a coding tree. ²¹ Inductive approach was used to derive themes and sub-themes from the advanced codes.

The quantitative data was analysed by frequency distribution and descriptive statistics. Mean (SD) was used to describe categorical variables (nominal) and count (percentage) was used to represent continuous variables (ratio). SPSS version 26.0 was used to analyse all randomized patient responses gathered through the intervention receipt questionnaire, the intervention contamination among control arm questionnaire and patient baseline and feedback data from electronic case record forms.

Results

The results of PE are presented using both frameworks with triangulation of quantitative and qualitative results. A summary of the PE findings is presented in Figure 2 using Realist framework of Context-Mechanisms-Outcomes.

Figure 2.

SPRINT INDIA process evaluation findings summary.

The various themes and sub-themes identified are presented in Table 2.

Table 2.

Qualitative themes and sub-themes with patient perspectives.

Themes and sub-themes	Description	Patient perspectives
I. Fidelity to Protocol Adherence Follow-up Coordinator support Challenges Administrative Support	There was a mix of SPRINT protocol adherence with training, delivery and follow-ups with administrative support. The issues were identified.	It said not to smoke, to take tablets on a regular basis, not to get agitated, and other minor details, as well as to take the tablets appropriately while travelling out of station – A Patient
		Many patients don’t like that we make so many repeated calls. Whereas there are cooperative patients, like we called them and asked about their wellbeing. What are the issues they are facing and all – A Centre Coordinator
		Now I have to pick him up at 5:30pm. Then I spend time with the baby girl or doing some household work. So I am busy for the whole day – A Patient
II. Fidelity to Intervention Feedback of intervention Adherence to intervention Effectiveness Challenges	The perspectives showed the appreciation of the three pronged intervention but the dilemmas presented issues in uptake.	For example, many of the patients, they quit their smoking and they stopped their alcohol in taking and they stopped their intake of more salt and they stopped in taking of cheese, butter, these kinds of food items. Definitely we have seen the change – A Centre Coordinator
		Sometimes meals was quite irregular. . .But now everything is on time, now I eat my lunch around 12:30 and dinner also I try to finish by 7 PM and I don’t eat anything in between. Sometimes I drink soup. . .In between only fruits. – A Patient
III. Impact of Intervention Added benefits by intervention Lifestyle changes Reintegration	The intervention helped not only the current health status of patients but caused unanticipated advantages.	Because one thing is, it will be very relevant in hospitals where you don’t have a dedicated stroke unit and people to educate the patients. So, if you deliver this as a package, in such centres, it will be very useful for the patients. – A Centre Investigator
		Yes, my family member also eat good diet and avoid the oily food as we used to eat non-veg a lot, now we came to know that we should not eat non-veg more. There is lot of cholesterol in the non-veg food. All the oily food is stopped now – A Patient
		Physical activities were the same earlier as it is now. Exercise has increased a bit now like I prefer walking back home rather than taking transport – A Patient
IV. Barriers Accessibility Literacy Affordability Staff-Patient communication Caregiver dependency Stroke awareness Impact of COVID Intervention	The stakeholders perspectives highlighted various issues. The range of barriers exposed a few known and some unexpected determinants.	Illiteracy, not living with the old patients and a basic mobile phone limits inclusion in the SPRINT – Centre Coordinator
		Yes, see it once or twice. . . I’ve had a lot of work to do, so I don’t have much time – A Patient
		Most of the patients tell me regarding their financial problems, due to financial problems they can’t able to buy their medicines. – Centre Investigator
		Whenever I receive a text message, I forward it to my son, who reads it. I’m an illiterate. – A Patient
		My mother’s phone isn’t a smartphone; it’s a regular phone that doesn’t support Tamil, which was an issue for her. – A Caregiver
		The study material that we have provided I think it should have been more concise. The length of the work book was too long, it should have been more concise and more to the point. – A CCC member in Focus group

Implementation of the trial

Reach

Eligibility evaluation was carried out on 5640 patients; out of them 1342 were excluded due to reasons like severe comorbidities (172), did not own smart phone (384), refused consent (759) and unavailable for study (27). As depicted in Figure 3, 231,781 SMS and videos were delivered to 2148 intervention arm patients. Acknowledgement of SMS and video receipt was received for only 38,545 (17.0%).

Figure 3.

Quantitative results of SPRINT INDIA process evaluation.

Many health professionals felt that few patients were unsuitable to participate due to illiteracy, cognitive difficulties, busy schedules and limited internet access. Few coordinators at high volume centres and those with a lack of infrastructure, had created their own check lists for restricting screening.

“. . .So due to some time constraints, we cannot screen all although I am trying to screen most of them.” – A coordinator

Effectiveness

Patients were able to adhere to instructions provided by coordinators. Patient follow-up by research staff, facilitated with arranging for doctors’ appointments. Help was also provided to control arm participants.

“Whatever exercise I was told to do in the book, I liked it, the way of eating that has been explained also I liked it and whatever explained about the fruit, which will keep my health good, it was all good.” – An intervention patient

“We spoke a lot about medication and they know what they have to take. We call them and tell them about medicines, do regular exercise and what you have to eat and all we tell them.” – A coordinator about control patient

Challenges

Patient disability, lack of caregiver support or inability to understand intervention were reported in following the protocol.

“She (patients’ daughter) told me about those messages (intervention), and forwards some of them. She showed it (videos from the intervention) to me a few times, but I had no idea what she was talking about.” – An intervention patient

Other challenges included the impact of COVID, completing online case record forms, completing follow-ups, patient misconceptions and inconsistent research staffing.

“During COVID time we actually plan the meetings. It helps the site to clear the queries and do the follow up with them. But some modifications were done. It actually did not work.” – A CCC member in FGD

“Every time we appoint a coordinator, we train them fully, provide study material and we give them online training or we go there and train the staff. Problem is that the coordinators are frequently changed.” – A CCC member in FGD

Figure 3 showcases an outline of the integrated quantitative results. Fidelity questionnaire data for capturing acknowledgement of SMS and video receipt from the intervention arm, reveals general decrease in viewing messages.

“Yes, like when a message that had already come used to come again, I used to tell my husband that I have already heard all this. And like you should tell me something else” – An intervention patient

Contamination

Intervention contamination questionnaire data for control patients from Figure 3, shows that no one acknowledged information of intervention. However, at one of the centres, a stroke patient who was ineligible to participate during the screening process, came to know of the intervention and requested inclusion.

“. . .it once happened that in ICU I have to randomized a patient so I brought attendant separately and explain them they receive the book, it was done confidentially but somehow next bed ICU patient came to know, who was also stroke patient and didn’t meet all criteria so I was not taking other patient so at that time they asking the administration she get the stroke book I also need stroke book. . .” A coordinator

Adoption and effectiveness of the intervention

Interventions facilitated patient and caregivers’ education and were a reminder to adhere to treatments. Most patients were satisfied with the three interventions. In the Intervention feedback questionnaire, out of 901 respondents, n = 415 (46.1%) found the experience of taking part in SPRINT INDIA Trial excellent, n = 434 (48.2%) said it was good, n = 44 (4.9%) average and only n = 8 (0.9%) found it poor.

The SMS proved beneficial for the intended population.

“See, one thing I gained from reading the SMS messages was that I was aware of what was going on. . .I was aware of the treatments that were taking place as well as their repercussions.” – A caregiver

Patients found the workbook activities engaging and motivating.

“Yes I liked the activities very much. I did not want to return back the book to them but it was said that I have to return it back after one year so I returned them back.” – A caregiver

“He (patient) used to take medicines irregularly before reading book but after reading the book, now he used to have meals at time and he used to say that give me food on time so that I can take medicines and do exercises on time.” – A caregiver

Calls from coordinator were a catalyst to make the changes as advised.

“. . .doctor said nothing and merely prescribed medications. . .coordinator simply explained everything in depth, demonstrated all exercises, and provided information about book”– An intervention caregiver

Challenges

Patient’s illiteracy, difficulty in follow-ups, dependency on caregivers and monotony of same videos and SMS.

“I don’t know how to use it, due to my poor literacy and poor education, and anyway, I cannot do anything about it.” – An intervention patient

“. . .they visited any local doctor in interiors, and he told them that your pressure is alright and why are you unnecessarily taking these medicine.” – A coordinator

Additionally there was a problem with accessing the videos which were due to a lack of a smart phone, poor internet connectivity or lack of resources to purchase a smart phone.

“internet issue and there is a information written there about the videos, so patient used to read that and give a missed call but mostly they haven’t open that link. . ..” – A coordinator

The number of returned workbooks is considerably low with completed activities (Figure 3) due to loss of workbooks, getting busy after recovery or caregiver dependency.

“In the beginning, I read that book daily but after I got busy with some other work, I stopped reading it” – An intervention patient

An investigator succinctly articulated that no value is attached to a free article.

“When you give anything free of cost, people don’t tend to ask details. . . I had some experience where the patient did not read the book, but some of his relatives saw the book and he was very interested, he read it thoroughly and he called me over phone sir this is a very good book, can I buy it?”

Impact and maintenance of the intervention

Interventions were found to be effective and facilitated in patient’s recovery, as a reminder of the post-discharge instructions and lifestyle changes.

“Yes madam,. . .they will know what monitoring should be done after a stroke like monitoring blood pressure, which medicines are there, such as how to take blood thinner drugs,.. It has a great impact on patient care on how to manage patient at home after stroke. . .” – A neutral health professional

Patients had easy access to consultants and laboratory investigations through coordinator communications. The study also provided a better working environment for the centre’s stroke team.

“If symptoms happen to them so we help them after consulting with the doctor, if they feel their follow ups is not possible and medication is going to finish so we provide them prescription medication after consultation with doctor.” – A coordinator

“Through this trial, we have got some additional hands in the stroke team and there is a good comradery. . .” – A center investigator

Patients were also able to make modification to diet and lifestyle and adhere to medications. The intervention helped with resuming the patient’s daily life.

“I always tried to do a little bit of what was in the book, I tried to go from five minutes to ten, ten to fifteen minutes, to see how much exercise I could. . .My book has benefited a lot.” – An intervention patient

Challenges

The challenge of illiteracy and lack of device to view intervention hampered in maintenance.

“. . .initially we thought that they are poor and they cannot purchase the medicine. . .but here each and everything is free. But because of their illiteracy. . .they neglect each and every thing we tell them.” – A coordinator

Discussion

Our findings support the assertion that SPRINT INDIA was a methodically executed study of a three-pronged intervention that balanced an innovative and culturally appropriate intervention with standard secondary stroke awareness. ¹⁴ In addition, PE followed predetermined protocol, which creatively harmonized the REAIM-Realist frameworks and triangulates the quantitative and qualitative data findings to indicate trial implementation and the facilitators and challenges.^17,18

Both the workbook and mHealth interventions have been researched in past, but their practical applicability has not been fully understood with the addition of new elements.^22,23 We have determined possible explanations wherefore we were unable to achieve the anticipated trial outcome using the Realist framework of Context-Mechanisms-Outcomes (as shown in Figure 2). ¹⁸

We discovered that intervention fidelity and adherence to the protocol were generally excellent across centres, barring the fact that availability of having a device to view intervention was only conceptually supported but not extensively implemented. Consistent with earlier studies, there are impediments to adoption of digital health when the device is shared or when there are low levels of technological literacy among stakeholders.^24,25 Following the consent, of the caregiver, patients with sub-acute stroke were randomized into trial without ensuring their education levels or their not possessing a personal device to view intervention. It was observed that over time, as the patient recuperated, the caregiver got back to their job or left the town causing termination of the intervention. Hence, it’s critical to accurately identify which patients qualify to derive advantage from intervention. ²⁶

Another issue which created bias was that the coordinators frequently spent time counselling (not specified in the protocol), to both the groups which may have caused inadvertent bias. The control and intervention patients and caregivers would seek the coordinator for guidance and request quick appointments with the doctor. As a result, cardiovascular protection improved in routine care, by virtue of participating in the trial. ²⁷ During the initial phase of patient care, caregivers and patients felt a need for counselling as they are busy adapting to the new grievous situation and cannot concentrate on the intervention. This observation is in line with other sub-acute stroke interventions. ²⁸

The patients who were following intervention were bored overtime with repeated material. Perhaps a step-up approach of delivering another set of messages, videos and workbook with more customized information would have increased the effectiveness. Alternatively, a large proportion of patients claimed to have used intervention without acknowledging receipt of SMSs and videos. This reemphasizes the difficulty to measure utilization of mHealth interventions accurately. ²⁴ In our case, SPRINT INDIA may have been used more frequently than reported.

We feel SPRINT INDIA PE is a robust evaluation conducted with purposive selection of 115 different stakeholders from 11 centres in 11 different languages along with focus group discussions. Context-Mechanisms-Outcomes design and theoretical frameworks, particularly centred on themes of Reach and Maintenance, were useful in integrating our findings. It offers fresh perspectives on the mHealth implementation process in stroke prevention demonstrating, how the intervention elements interacted and were in synchrony with each another. All the same, some limitations need to be acknowledged. PE was supposed to be done parallel to data collection and conducted in person at all centres, but this was not possible due to termination of trial and pandemic. Purposive selective sampling was compromised at a few centres as invited caretakers/patients had issues coming to the hospital during pandemic. We are unable to fully explain which intervention is more popular or impactful; which is a common drawback of complex interventions. ²⁹

In hindsight, additional in-depth exploratory research with qualitative interviews of stakeholders during pilot trial may have identified finer implementation and dosing measures to promote efficacy of intervention. However, we have given several unanticipated consequences that affected the impact and explanation of potential imminent pathways which were not clarified by the quantitative results.

Conclusion

According to our research, SPRINT INDIA should be developed further, as it met a critical need in providing stroke awareness in India because this type of educational intervention is unavailable and inaccessible. Future iterations of the trial must be more tailored to the participant’s environment. The workbook might be divided into two categories: one for literate people, which might contain more illustrations, and one for illiterate people, which might have more pictures and less text. The mHealth component can be made adaptive in a similar way; SMSs and videos might be freely provided to people who have personal smart phones, while people without smart devices might just receive SMSs.

Moreover, provision of upgraded material over time will sustain intervention and increase its effectiveness. A stringent adherence to protocol inclusion criteria and timing of inclusion would help in its implementation. Involvement of community-based service implementation by Accredited Social Health Activist workers and primary health care centres can be considered for marginalized underdeveloped populations. Evaluation results are crucial when deciding how to scale up and adapt the programme to different contexts by policymakers and researchers given the phenomenal penetration of mobile devices into population.