Diversion of Diacetylmorphine in Heroin-Assisted Treatment: A Focus Group Study among Patients and Treatment Providers Open Access

Heroin-assisted treatment (HAT) was introduced in Switzerland in 1994 as a response to emerging open drug scenes, increasing overdose deaths, and increasing prevalence of blood-borne diseases [1]. HAT involves the prescription of injectable diacetylmorphine (DAM; pharmaceutical heroin) and of oral DAM tablets to patients with severe opioid use disorder (OUD) who respond poorly to conventional opioid agonist treatment (OAT) [2]. Possible routes of DAM administration include intravenous, intramuscular, and oral, with intranasal administration currently being evaluated [3]. Since its introduction in Switzerland, several European countries and Canada have implemented HAT, which has been shown to effectively retain patients in treatment who previously responded poorly to conventional OAT (e.g., methadone) [4]. Furthermore, long-term HAT is associated with improved physical health, mental health, social health, labour-market integration, as well as a reduction in criminal offending [5‒7].

Compared to conventional OAT, HAT is more strictly regulated in Switzerland. DAM is only dispensed in specialised treatment centres and is typically administered under supervision due to safety concerns. Unlike slow-acting opioid agonists, patients receive take-home DAM only if they meet stringent criteria defined in the Swiss Narcotics Act: patients must have been in HAT for at least 6 months and must be physically, mentally, and socially “stable.” Notably, the legislation does not further define the term “stable.” Additionally, a patient’s last two urine screens must be negative for illicit substances and the risk of misuse estimated by the prescribing physician must be low [8]. These regulations are in place to ensure patient safety and minimise the risk of diversion and misuse. Whereas before the COVID-19 pandemic, take-home prescriptions in HAT were only allowed for a total of 2 days, the regulations were temporarily relaxed during the pandemic and then permanently adapted in 2023 to allow for a total of 7 days [9]. Internationally, Switzerland is an exception in this regard as no other countries allow HAT take-home prescriptions.

In the context of prescription drugs, diversion is defined as any intentional act that results in transferring a drug product from lawful to unlawful distribution or possession [10]. Diversion is associated with a range of individual and societal harms, such as overdose, accidental ingestion, and high societal cost due to criminal justice procedures, loss of productivity, and treatment of medical complications [11‒14]. Generally, medications may be diverted for economic reasons (i.e., to sell on the black market) or for misuse (e.g., using DAM through a non-intended route of administration) [15]. However, the literature on diversion in HAT is scarce, with no recent studies investigating this phenomenon directly.

Indeed, the occurrence of diversion, underlying reasons for diversion, and the consequences of diversion in HAT have not yet been investigated. Furthermore, there is no general policy in Switzerland on how to deal with diversion incidents. Attitudes and assumptions from the treatment provider’s perspective regarding diversion are also unclear and not available in medical literature. The present study aimed to fill this gap of knowledge by investigating the reasons for, experiences with, and consequences of DAM diversion from the perspective of patients and providers.

We used a combination of a deductive and an inductive approach to hold two focus group discussions. The first focus group included treatment providers, while the second included patients receiving HAT. The study was conducted in the outpatient HAT centre of the University Psychiatric Clinics Basel, Switzerland. The clinic is open from 9:00 to 11:30 and from 14:30 to 17:00. The maximum number of HAT dispensings a patient can receive in a single day is two: one in the morning and one in the afternoon. At the time of the study, 160 patients received HAT at the study site. Treatment providers and patients were recruited by convenience sampling by the responsible study physician, who directly approached potential participants for both groups. In order to increase diversity and representativeness, we aimed to include members of all professions involved in treatment in the provider group, and patients with varying treatment duration, age, sex, route of administration (i.e., injected, oral, nasal), and take-home privileges. Inclusion criteria for patients were willingness to participate and ability to provide written informed consent. No exclusion criteria were defined. Twelve patients and 13 treatment providers participated in the focus groups.

Prior to convening the first group, the research team developed a topic guide for the provider group in order to initiate discussion in case this became necessary to cover all previously identified relevant aspects. This topic guide was based on the literature on diversion in OAT in general, and on the clinical experience of the authors with diversion in HAT specifically. It included open questions on HAT in general, HAT medication dispensing, characteristics of HAT diversion, underlying motives, consequences, and suggestions on improvements. The themes that were identified from the analysis of the treatment provider group were then used to adapt the topic guide for the subsequent patient group. No distinction was made between DAM tablets and injectable DAM in the topic guide.

The first focus group included nurses (n = 6), social workers (n = 4), psychologists (n = 2), and a psychiatrist (n = 1). Seven participants were identified as female, six as male, with a mean age of 41 years (range 25–64) and a mean duration of employment in the treatment centre of 8.5 years (range 0.1–28). The second focus group included 12 patients, 5 of whom were identified as female and 7 as male. The mean age was 47.3 years (range 20–58) and the mean duration of receiving HAT was 10.5 years (range 0.2–23.4). Four patients had a prescription for injectable DAM only, 5 received a combination of injectable DAM plus oral DAM tablets, 2 received intranasal DAM only, and 1 received a combination of intranasal plus oral DAM tablets. Two patients had a take-home prescription of DAM and 10 received additional opioids in the form of take-home slow-release oral morphine tablets.

Focus groups were audio-recorded and moderated by researchers not involved in treatment (E.S., Z.W.) to reduce any potential bias in patients’ discussions due to potential fear of negative consequences for treatment. E.S. is a psychiatrist and senior physician in a psychiatric university clinic and Z.W. was a master’s student in medicine. Both researchers were internally trained and supervised by senior research team members (K.M.D., M.M., M.V.) with extensive experience in treating patients with OUD and in conducting qualitative research. No research team members involved in treatment participated in the focus groups.

E.S. was the main moderator of the sessions and Z.W. mainly took field notes. Both researchers introduced themselves and their respective roles in the study at the outset of the groups and informed patients that a similar focus group had been previously conducted with staff. Moderators explained that the aim of the groups was to gain knowledge of participants’ views and experiences concerning diversion in HAT. Furthermore, patients were ensured that anything discussed in the focus group would be anonymised. Focus groups were conducted in person during October and December 2022 and held mainly in standard German or in the Swiss German dialect. As there is no written standard for the Swiss German dialect, audio recordings were transcribed into standard German by E.S. Moreover, any information that could identify a participant was removed by E.S. at this stage. The duration of the treatment provider focus group was 65 min, while that of the patient focus group was 55 min. No time limit was defined prior to the focus groups and discussions continued until no new topics emerged. Participants of the patient group received monetary compensation, and both groups were provided snacks and drinks.

The emerging corpora from both focus groups were analysed using Mayring’s qualitative content analysis framework [15]. This well-established method of qualitative text analysis consists of an inductive approach, in which data are assessed without presumptions or theories. E.S. and Z.W. separately applied codes after defining selection criteria for codes. MAXQDA 2020 (VERBI Software GmbH, Berlin, Germany) was used for coding. Codes were compared and used to build overarching themes and categories. Any differences and uncertainties regarding the application of codes and the abstraction-level of themes and categories were resolved in meetings with the project supervisor (M.V.). Quotes selected for this article were translated verbatim into English by M.V. and translated back to German by M.M. to ensure semantic consistency. We aimed to remain as close as possible to the original citation and thus largely refrained from rephrasing to fit the English language style. Within the results section, we refer to additional quotes that can be used to provide more context (online suppl. quotes; for all online suppl. material, see https://doi.org/10.1159/000545162).

Several themes emerged from the focus group discussions, summarised in Figure 1. Each theme contained several categories. While providers perceived diversion attempts as common, both patients and providers reported that only a few patients are involved in diversion.

Providers as well as patients speculated that some patients who divert their medication received higher than needed doses.

While it is easy to divert excessive, unneeded doses for patients who receive take-home, those with supervised administration develop different techniques for sneaking medication out of the treatment centre. Treatment providers described such observed attempts at diversion and, while confident that they could sometimes recognise imminent attempts, they also mentioned the difficulty of closely supervising several patients simultaneously (online suppl. Quote 1). Providers also distinguished between spontaneous and premeditated diversion.

The use of other substances in the context of comorbid substance use disorders emerged as a main motive suspected by patients and providers. Both groups mentioned that selling diverted DAM could help pay for other substances, particularly cocaine, among patients with cocaine use disorder. Patients thought that diverting DAM was more justified if the money was then used to finance other substance uses, as opposed to any other motives.

Moreover, patients and treatment providers also mentioned that DAM is still rare on the black market in Switzerland. It, therefore, has a high price which increases the incentive to sell it (online suppl. Quote 2, 3).

An additional motive for diverting DAM mentioned among patients was to gain autonomy with their DAM use, particularly as it relates to the route of administration (online suppl. Quote 4). The clinic only allows the use of intravenous, intramuscular, oral and, only since 2020, intranasal liquid DAM. Therefore, for patients wishing to use DAM through any other route of administration, there is a stronger motivation for diversion.

Providers and patients also brought up that some may prefer to take DAM alone at home, as opposed to in a room with many other patients under medical supervision (online suppl. Quote 5). Patients expressed the wish to decide autonomously on how much, how frequently, when, and where to take their DAM.

Although the limited opening hours and available time slots for dispensing helped structure the day for some patients, patients also viewed these as confining, leading them to opt for diversion of DAM in order to tailor intake to their own needs.

One patient reported the necessity to postpone intake to be functionally “sober” at a fixed timepoint. He had diverted DAM in the past because he had to attend a family celebration after dispensing times when the clinic was closed.

Providers described the effect that DAM may have on patients’ feelings. Some patients may use DAM for emotional regulation, resulting in an inclination to take multiple smaller, self-determined doses as needed, rather than higher doses twice a day during opening hours (online suppl. Quote 6). Patients also described the desire to combine different substances to modulate the subjective effects.

Providers speculated that medication-sharing with peers who are not registered for treatment may constitute another potentially relevant motivation for diversion.

Patients noted that sharing or selling DAM may result from third persons’ wish for confidentiality. Some people may hesitate to enter treatment because they do not want their use to become known to the health insurance and thus use diverted DAM (online suppl. Quote 7).

Treatment providers also described that the entry criteria for HAT impede access to HAT. People with high-risk opioid use (HROU) not meeting the criteria would look for DAM on the black market, incentivising diversion by those with access.

Patients and treatment providers discussed several negative consequences resulting from diversion, ranging from health emergencies and security aspects to treatment policies.

Three categories of increased overdose risk emerged: within the treatment centre, outside the treatment centre, and among opioid-naïve individuals taking diverted DAM. Providers explained that patients who were caught diverting or selling tablets were subsequently sanctioned to receive their prescribed amount as crushed tablets dissolved in water for supervised intake. However, the prescribed amount given as tablets could exceed a patient’s opioid tolerance, especially if the patient has been diverting parts of their prescription in the past and has, therefore, been taking less than what was prescribed to them.

The risk of overdose outside the treatment centre was noted by patients who described the dangers of using DAM in private.

Another concern of patients and providers was the use of DAM by opioid-naïve individuals. Both groups deemed the use of diverted DAM a high risk for fatal outcomes. DAM tablets are only available in 200 mg doses, which can be a lethal dose for people without opioid tolerance (online suppl. Quote 8–10).

Providers noted a substantial negative impact on the therapeutic relationship with patients. Confronting patients when diversion is noticed may lead to conflicts. Nurses expressed their feelings of helplessness, who are often exposed to hostility and insults as they are the ones handing out the medication to the patients.

Treatment providers described the difficulty of maintaining a therapeutic relationship with patients they have called out for diversion. In the clinic, individual treatment providers are appointed as the first point of contact for a number of patients. They, therefore, spend more time with these patients compared to with other patients. They can then become more hesitant in confronting their patients about diversion as this may jeopardise their relationship.

Nurses also mentioned the ethical and relational difficulties that arise from controlling whether patients have really swallowed the oral medication. They described a perceived obligation to control medication intake and experienced this as stressful (online suppl. Quote 11). Other professions involved, such as social workers and doctors, described the procedures around diversion as less stressful than nurses (online suppl. Quote 12, 13). In contrast to the strongly felt impact on the relationship described by providers, patients mostly felt that other than transient dispensing of dissolved tablets under supervision, there was no sustained negative impact on treatment when treatment providers identified diversion (online suppl. Quote 14).

However, some patients mentioned the fear that diversion may lead to stigmatisation among treatment providers. They expressed frustration at being collectively mistrusted because of isolated instances of diversion.

Patients described the fear that the misuse and selling of take-home doses may lead to more restrictive regulations, thus reducing patient autonomy (online suppl. Quote 15).

Another patients’ concern was the fear of diversion, leading to the closure of the treatment centre. As there are only two centres prescribing DAM in the region, diverted DAM on the black market can be traced back to these clinics.

Both groups also described the potential positive effects of DAM being obtained by opioid-dependent individuals on the black market.

Patients and treatment providers described that diverting could potentially recruit new patients into treatment. Some individuals with OUD acquiring DAM from the streets may decide to enter treatment because of positive experience with the DAM and reimbursement by health insurance.

Providers noted that if DAM is sold on the black market, individuals are assured that they are receiving DAM which has not been mixed or diluted with other substances, and there may be fewer negative consequences (e.g., overdose) as a result.

Patients and providers suggested several strategies to prevent or reduce diversion, such as offering different and more preferred routes of administration for DAM as well as the flexibility to switch or combine prescribed routes (online suppl. Quote 16).

Take-home dosing as another way to reduce diversion was brought up. Patients described the reduced prices for DAM on the black market in settings where more liberal take-home policies were practiced (online suppl. Quote 17), probably because of decreased demand for diverted DAM. They also mentioned the restrictive nature of supervised dispensing, potentially preventing them from incorporating treatment into their everyday life. Indeed, if the clinic is open at 8:30 and closes at 17:00, this makes it difficult for individuals with a 9-to-5 job to come to the clinic and be in treatment.

Having said that, 1 patient mentioned that he perceived the given timeslots for dispensing in the HAT centre as positive, helping him to structure his daily life (online suppl. Quote 18).

Providers also noted the difficulties in determining the “stability” of individuals, which is an inclusion criterion for HAT as required by law (online suppl. Quote 19). They also mentioned that the restrictive nature of take-home practices in the centre contributed to diversion (online suppl. Quote 20).

Patients and providers suggested infrastructural changes to the treatment centre, pointing out that entry criteria for HAT divide individuals with HROU into those who receive DAM and those who are forced to buy it on the black market if they want to get it (online suppl. Quote 21). Providers also pointed out the need for more personnel to allow thorough supervision during dispensing.

Patients also mentioned that an adjustment in opening hours could be helpful, to allow more flexibility in the timing of DAM use. This overlapped thematically with the category of autonomy in the motives theme.

Patients and treatment providers noted the necessity of expanding treatment strategies and interventions. Providers stated that understanding patients and their individual situation would be another important aspect in reducing diversion and suggested psychosocial strategies like motivational enhancement (online suppl. Quote 22). The patients also suggested that cocaine prescription could be helpful in reducing diversion.

This qualitative study investigated the phenomenon of diversion in a Swiss HAT centre. Overall, the two focus groups reported diversion of DAM tablets and injectable DAM seems to be uncommon. The two groups acknowledged that diversion occurs for reasons related to infrastructure, regulatory restrictions, limitations in treatment, and non-regulated substance use, particularly cocaine. Of note, it is in part the same factors that promote diversion that constitute barriers to treatment entry.

A recurring motive for diversion was the desire for more autonomy, relating to time of day and mode of medication use. Take-home prescriptions allow patients to take the medication according to the requirements of their daily lives, while also respecting the fact that the practice of injecting is often an intimate process that can potentially be disconcerting in supervised settings [16]. The importance of take-home doses stems from HAT being an intensive treatment setting, for which patients must attend several times daily if no take-home doses are prescribed. In a study among individuals receiving injectable DAM or hydromorphone in Vancouver, Canada, patients prioritised attributes that would empower them with greater autonomy [17]. Among the attributes, take-home doses were the most prioritised feature overall. In Switzerland, before any take-home of DAM can be prescribed, the legal regulations require patients to be in treatment for 6 months, whereas there are no such limitations for conventional OAT [8]. However, for patients deemed “stable,” providers can apply for exceptions from the Federal Office of Public Health. Notably, legislation does not give a precise definition for “stability” in the context of HAT, making this criterium potentially unhelpful for clinical practice. These regulations also disregard the fact that some patients may require take-home prescriptions in the first place in order to “stabilise.” With the restriction on take-home medication, patients are forced to attend the clinic at specific times during the day, making it difficult to maintain regular activities such as employment or educational curriculums, which constitute important indicators of “stability” [18]. This constitutes an important treatment barrier and is a point that has been raised repeatedly by patients in the present study, as well as in other studies from settings in which take-home prescriptions are prohibited [19]. Particularly, individuals working regular hours (9:00–17:00) with HROU might hesitate to enter HAT. Facilitating take-home access could be instrumental in recruiting these individuals for treatment. In the aforementioned Canadian study, patients reported that being able to access the treatment centre on their own schedule was another important attribute that could improve treatment engagement and retention [17].

Diversion out of a desire for more autonomy can directly be attributed to the strict legal framework in which Swiss HAT operates but is likely generalisable to every treatment setting with restrictions on take-home prescriptions (in particular oral OAT). As for patients who are already in treatment, providers agreed that it is challenging to decide who is eligible (i.e., “stable” enough) for take-home doses. This decision remains a conundrum that can only be solved by a strong, good-quality therapeutic relationship. In turn, enforcing strict legal criteria entails the risk of damaging the therapeutic relationship, which has previously been observed [16], when decisions are perceived as unfair and stigmatising by patients. Rather than calling out patients who divert or ignoring diversion attempts, understanding the impact of barriers to take-home prescription and compassionately engaging with patients in order to formulate a plan to address these barriers is a better alternative. For patients who respond poorly to conventional OAT, the change to HAT as the most effective treatment should be facilitated. Eliminating arbitrary limitations to treatment access will reduce the incentive to divert medication. Likewise, a reform of the rather vaguely worded legal criteria for take-home prescriptions may be helpful. The current regulation using unclear and potentially stigmatising terms such as “stability” is prone to arbitrary decision-making by the prescribing physicians and contributes to asymmetry in the patient-provider relationship. On a positive note, however, the vague terminology also allows individual treatment choices and may even result in more take-home prescriptions than would be the case with too narrowly defined criteria.

Equally, patients brought up the wish to use different routes of administration and to be able to adapt their use to their daily needs or life circumstances. Indeed, patients noted a reduced necessity to divert DAM tablets following the introduction of intranasal administrations at the study site in the setting of a naturalistic prospective cohort study [3]. In contrast, inhalable DAM as another route of administration is not yet available in Switzerland, despite having been shown to be a viable treatment option in the Netherlands [6]. As a result, patients who prefer to smoke DAM tend to divert their medication or revert to the black market. Efforts to reduce diversion should, therefore, focus on individualised interventions and addressing patient preferences, particularly through diversifying prescribable routes of administration.

Some issues brought up by the focus groups can be addressed more easily than others. Both groups described that patients may ask for dose increases beyond what is needed in order to divert the surplus dose. For dose-finding in HAT, treating physicians often rely on the reports of their patients’ concerning symptoms, such as craving or subclinical withdrawal. Furthermore, the wish for independence from restrictions imposed by treatment structures such as limited opening hours and a maximum of two daily dispensing seem to be an important driver. Patient autonomy can be increased by take-home prescriptions [17], but no satisfying solution yet exists for patients for which supervised dispensing is necessary.

Patients and providers alike noted the substantial impact of diversion on treatment. On the patient-side, diversion can reduce treatment quality (e.g., by patients being forced to use a different route of administration such as swallowing dissolved tablets) and impair patients’ ability to pursue other life activities (e.g., daily supervised dispensing after revoking take-home privileges). However, treatment providers, in particular nursing staff, were more personally impacted by the effects of diversion. Indeed, they are primarily responsible for repeatedly reminding patients against diversion and for enforcing the clinic rules as it relates to diversion, which was described as stressful. In the nurses’ perception, sanctions like dissolving tablets for supervised intake exposed them to insults and endangered the therapeutic relationship. One nurse even stated to have refrained from confronting patients suspected of diversion to avoid undermining their relationship. This underlines the importance of the therapeutic relationship and trust between providers and patients. They are key factors in promoting healthcare access and retaining vulnerable populations in treatment [20]. Instead of directly confronting or ignoring incidents, providers could employ alternative approaches such as compassionate case reviews and individualised solutions based on patients’ needs.

On a public health level, diversion in HAT might, in an unlikely worst-case scenario, endanger opioid-naïve individuals due to overdose risks. Limited data are available to assess these risks, but a recent study investigated the effects of prolonged DAM take-home and found no increase in emergency hospitalisations or incarcerations [21]. Concurrently, diversion in HAT may also lead to a recruitment effect, as noted by providers and confirmed by a participant from the patient group. Diverted DAM emerging on the black market might attract individuals with HROU as the higher purity of the substance may also increase the subjective effects after administration. This phenomenon was previously described in Switzerland, when the number of patients in treatment increased following the loosening of take-home prescription criteria during the COVID-19 pandemic [22]. Furthermore, providers speculated that compassionate sharing with peers might be a driving factor of diversion in HAT. Previous literature has described this motive in the context of the North American overdose crisis, in which patients risked the termination of their OAT by diverting medication to be able to supply opioid agonists to others [23]. Hence, in order to reduce diversion, treatment coverage in OAT needs to be improved. For example, this can be done by reducing access barriers and making OAT more attractive to the target population.

People with HROU have a high incentive to buy DAM on the black market because there are still significant barriers to treatment. The demand could be reduced by abolishing the restrictive entry criteria to HAT, which are largely derived from the criteria of Swiss HAT studies from the 1990s and have never been systematically evaluated. Patients not meeting all entry criteria may also benefit from this treatment, consistent with findings showing that HAT was effective in patients without prior OAT [24].

A strong motive for diversion was co-use with other substances, particularly cocaine. Both groups acknowledged the need for novel therapeutic interventions of cocaine use disorder to address diversion in HAT clinics. On the one hand, increased psychosocial treatment can be offered by increasing centre treatment providers, thereby expanding resources for individual and group therapies. On the other hand, no medication has yet been approved for the treatment of cocaine use disorder. With that being said, a meta-analysis of 38 randomised clinical trials has shown that prescription psychostimulants such as methylphenidate or modafinil can effectively promote abstinence [25, 26].

Our study has several limitations. We used convenience sampling and aimed to include patients and staff with diverse characteristics and backgrounds. Our participants may, therefore, not be representative of HAT in Switzerland and even more so in other countries with stricter regulations. Despite providing clarification that the content of the groups would not have any consequences for individual treatments, and the fact that moderators were not involved in any treatments, the use of focus groups as opposed to in-depth interviews may have influenced participants’ statements and led to socially desired answers. In-depth interviews could be advantageous in this respect and help identify additional themes and categories. Furthermore, we held the groups within the facilities of the HAT centre. Although we aimed to provide a setting in which patients felt safe to openly discuss diversion, we cannot rule out that the location might have impacted the results.

Diversion in HAT is a rare but challenging clinical phenomenon driven by the wish for more autonomy, concurrent substance use, and existing treatment barriers. In clinical practice, nursing staff were responsible for enforcing rules and, therefore, felt the greatest impact on their therapeutic relationships with patients. Responding to every patient’s needs and preferences in terms of opioid effect profile, route of administration, individual schedule, and social life, as well as improving treatment of concurrent cocaine use disorder, could greatly contribute to a reduction of DAM diversion. The Swiss legal regulations for take-home prescriptions appear unsatisfactory from both a patient and treatment provider perspective. This may actually be contributing to increased DAM diversion. Reforming these regulations, combined with tailored treatment strategies for severe OUD, is essential for reducing DAM diversion and enhancing overall effectiveness of HAT. Future research should investigate the effects of personalised approaches (e.g., more diverse route of administration and prescription scheduling) and regulatory changes (e.g., more flexible take-home prescriptions), as well as develop new strategies, to better address the needs of patients with concurrent substance use disorders.

This study was approved by the Responsible Ethics Committee (Ethikkommission Nordwest und Zentralschweiz, project-ID: 2022-01171). Study procedures were conducted according to the Declaration of Helsinki. All patients provided written informed consent before study participation. Participation was not remunerated.

The authors have no conflicts of interest to declare.

No funding was received for this study.

Elisabeth Strickler and Zacharias Wicki collected the data and conducted analysis. Maximilian Meyer and Marc Vogel designed the study and prepared the manuscript. Elisabeth Strickler, Jean N. Westenberg, Johannes Strasser, Undine E. Lang, and Kenneth M. Dürsteler revised the manuscript and provided substantial input. Undine E. Lang and Marc Vogel supervised the project and provided resources. All authors approved the final manuscript.

Maximilian Meyer and Marc Vogel contributed equally to this work.

The topic guide and data that support the findings of this study are not publicly available due to them containing information that could compromise the privacy of research participants but are available from the corresponding author, M.V.