The results of the process evaluation are described below, structured according to the process evaluation questions as presented in Table 1.
Fidelity (Q1)
All steps of the home visitation programme were largely implemented as planned. Exceptions are PNs who changed or omitted questions in the CGA, no or delayed post-discussion between the GP and PN, and no or inappropriately formulated care and treatment plans. Both PNs and GPs struggled most with how to arrange long-term monitoring of older people considering the limited number of hours PNs could dedicate to care for older people. Besides this, each PN developed his/her own routine in performing the different steps.
Change in mindset (Q2a)
The shift from reactively to proactively approaching older people in a structured and comprehensive way was evident for PNs, as well as for most GPs. One PN mentioned that GPs were not used to approaching older people in a proactive way. They usually offer care and/or treatment upon request, whereas PNs are more familiar with delivering preventive care.
At the end of the intervention period, all general practices intended to continue with the home visitation programme, but the proactive versus a more reactive approach posed a dilemma for half of the practices. Proactively visiting all older people (75+) allows for primary prevention of problems but is a huge time investment for general practices. Purposefully visiting older people who show signs of decline would be more feasible, but this is at the expense of detecting problems which could have been prevented if addressed earlier.
Interdisciplinary collaboration (Q2b)
Due to the home visitation programme, PNs’ extended their network of professionals or disciplines involved in care for older people and they used their network to a greater extent. The number of referrals to secondary care was limited and the majority of contacts took place within primary care, for instance with home care organisations, physiotherapists, and occupational therapists. Several PNs indicated that they only had few contacts with other care professionals since not that many problems had been detected. Collaboration with other professionals was facilitated when they were located nearby, preferably in the same building.
In three general practices, multidisciplinary meetings took place on a structural base (e.g., once a month) already before the start of this study, and older people were discussed in these meetings from time to time as well. In other general practices, no structural meetings were organised yet with disciplines outside the general practice, mostly because no or only few complex problems were detected among older people. Three general practices had concrete plans to organise multidisciplinary meetings in the near future. Others were not convinced of their added value compared to existing meetings or contacts with other disciplines on an individual base.
Decision-aids to support decision-making (Q2c)
Half of the PNs used the service map made available by the project team and considered it useful. Others had not encountered any situations in which the service map could have been helpful.
Both PNs and GPs were positive about the [G]OLD-instrument. Its extensiveness offered a comprehensive overview of the older person’s health and well-being, yet the instrument was time-consuming to administer. Consequently, several PNs and GPs sought for a balance in restricting the time investment without losing important content (e.g., application to a limited group of older people).
Clinical information system (Q2d)
Only in the region `Parkstad’, the digital system for registering the findings from the CGA and the care and treatment plan was finished at the start of the intervention period, although PNs could not yet register the results of monitoring and follow-up of older people over time. Initially, PNs considered it time-consuming and double work.
Dose delivered (completeness) (Q3)
The dose delivered is illustrated in Figure 2. In total, 590 participants were visited at home by the PN for a CGA between July 2010 and September 2011. An underestimation of the actual number of post-discussions is likely, since not all PNs consistently filled-out the registration form. The percentage of formulated care and treatment plans per PN varied widely from 4.0% to 95.2%. In case PNs did not use the official format, we did not count them as care and treatment plans. Finally, we had no valid details per older person of their follow-up within the chain of care as PNs could not easily differentiate between follow-up contacts on behalf of the home visitation programme and other contacts with the general practice within the 18-month period. Nonetheless, the follow-up process might have been suboptimal in several cases, as some PNs experienced time constraints and/or they did not have a concrete plan for monitoring and follow-up (this was not provided as part of the intervention protocol).
Dose received (exposure) (Q4)
According to the forms filled-out by PNs (n = 384, 65.1%), of the 229 older people who received follow-up actions or advice, 67.7% complied `always’ or `most of the time’, while 10.5% complied `rarely’ or `never’. Six older people admitted during the interview that they did not comply with a specific advice given by the PN.
Half of the PNs noticed that they often came across older people who did not want any follow-up action(s) in the first place. Especially mental problems were difficult to deal with. Often older people agreed to undertake actions when it was already too late. Some PNs struggled with how to deal with these older people and how to find a balance between respecting the older person’s wishes and maintaining contact to try to achieve the desired actions over time.
Dose received (satisfaction) (Q5-Q7)
Practice nurses (Q5)
PNs liked the performance of the home visits because of the ability to get to know the older person and to offer help or advice. The home visit lowers the threshold for older people to contact the general practice should problems arise in the future. PNs were in favour of a home visit instead of a consultation in the general practice, as it offered a more objective picture of the older person’s functioning. At the general practice level, the home visitation programme resulted in more attention for older people in general and closer collaboration between the GP and PN in organising care for older people. The majority of the PNs did not like the administrative work. Moreover, opinions of PNs diverged regarding the added value of the care and treatment plan over and above the registration of follow-up actions in the GP s Information System. In general, PNs evaluated the preparatory training rather positively (M = 7.64, SD = 0.50 for the first day vs. M = 6.64. SD = 0.92 for the second day).
General practitioners (Q5)
Half of the GPs liked most that they obtained a comprehensive and complete picture of older people s functioning and the social network surrounding older people. As a result, GPs considered the home visits useful, because it offered them additional information which might be valuable for future reference. At the general practice level, the home visitation programme had offered a starting point and useful tools for organising care for older people. Seven GPs mentioned that, against their expectations, no or only few previously unknown problems were detected. Furthermore, the older people that were out of the picture according to the GP and that would therefore particularly benefit from the home visit often did not consent to participate.
Older people (Q6)
Although the majority had no specific expectations about the home visit, three older people were hesitant about the purpose of the home visit at first. All older people were very satisfied with the home visit and afterwards, they had a good feeling about it or they emphasised that it had been interesting. They were very positive about the PN and felt that they could discuss everything with him/her. The home visit was neither too short nor too long and they felt that everything that they considered to be important was discussed. The questions asked as part of the CGA were not difficult to understand, impolite, awkward, or strange. Older people liked the ability to talk about different things, the unexpected attention from their general practice, and the fact that they now had a familiar face in the general practice. One person indicated that these visits tend to go towards too extensive meddling with other people s affairs, especially among older people who are doing relatively well and do not have a specific request for help.
Benefits to older people (Q7)
According to the forms filled-out by PNs (n = 394, 66.8%), for 29.9% of the older people follow-up actions, referral or advice had been `very much’ or `somewhat’ beneficial, while the remaining 70.1% of the older people benefitted `a little’ or `not at all’. Most PNs indicated that the majority of older people experienced few benefits, because no problems had been detected or only problems that could be addressed easily or problems that had already been taken care of. As recognised by a PN, a GP and confirmed by one older person, sometimes older people did not optimally benefit from the home visit because they to some extent hold a façade of normalcy. Finally, one GP commented that the home visits are less useful for older people without a specific request for help as they often do not want to undertake action.
Six older people mentioned that the home visit had been useful for them, mostly because it lowered a threshold to discuss matters for which they do not easily contact the general practice themselves. Others believed that you are just old and there is nothing that can be done about that, and that certain problems (e.g., loneliness) cannot be solved.
Reach (Q8-Q12)
Of the 1,972 eligible older people (≥75 years) approached, 36.6% consented to participate (see Figure 2). Mean age of participants was 80.6 years (SD = 4.26; range: 74.4-95.4) and 56.0% (n = 972) were female. Participants were significantly younger compared to non-participants (M = 81.2, SD = 4.39) (p = 0.004) and men were 1.43 times more likely to participate than women. Participants who dropped out (27.6%, n = 199) during the 18-month follow-up period (see Figure 2), received usual care if necessary. Drop-outs were significantly older (M = 81.1, SD = 4.76) compared to those who continued participation (M = 80.3, SD = 3.95), (p =0.027) and women were equally likely to drop-out as men (OR = 0.99). As Figure 2 shows, for only 38.0% of the 721 participants all steps of the [G]OLD-protocol up to the follow-up process were completed according to the registration forms filled-out by PNs.
Nearly all PNs and several GPs believed that they had missed the older people who would particularly benefit from the home visitation programme, since participants were the relatively healthy older people, those for whom care was already arranged quite well, or the ones who often visit the practice. They felt that due to the informed consent procedure of the trial, people who are not doing well or the more frail older people are suspicious about the consequences of participation.
Recruitment (Q13)
All general practices in the regions `Maastricht-Heuvelland’ and `Parkstad’ were informed about the [G]OLD-project by means of a letter from the primary healthcare organisation of their region, followed by information sessions and practice visits for those interested to participate. Non-responders were contacted by phone to inquire about their willingness to participate in the control group. Since insufficient general practices agreed to participate in the control group, jeopardising the continuation of the trial, the recruitment of control practices was extended to another region (`Midden-Limburg’). In total, 188 general practices from three regions were approached for participation and 24 general practices consented to participate (12.8%). Thirteen general practices were included in the intervention group (7 from the region `Parkstad’ and 6 from the region `Maastricht-Heuvelland’) and 11 general practices in the control group (2 from the region `Parkstad’ and 9 from the region `Midden-Limburg’).
Older people were approached for participation by means of an information letter and consent form. In the intervention group, those who did not return the signed consent form within two weeks were contacted by phone once to inquire whether they received the information letter. One postal reminder was send to older people who could not be contacted by phone. Due to the substantial time investment of calling older people, non-responders in the control group only received a postal reminder.
Context (Q14-Q21)
Barriers and facilitators for implementation (Q14)
Most of the barriers experienced by PNs during implementation were related to logistical difficulties in planning the different steps of the home visitation programme alongside other daily work. Especially the introduction of a new disease management programme for cardiovascular risk management (CVRM) by the primary care organisations during the intervention period posed challenges to PNs and GPs, causing several PNs to invest less time in the home visitation programme than planned. Finally, barriers for continuing the home visitation programme over time were the lack of an adequate reimbursement by health insurers of the costs of care for older people and the overall time investment of the home visitation programme (total time investment from preparation of the home visit to formulating the care and treatment plan was on average 85 minutes per older person).
A facilitator for implementation according to several PNs was gaining routine in efficiently planning and executing the different steps of the home visitation programme. Moreover, some PNs expressed the need for regular meetings with other PNs to exchange experiences or the ability to consult an expert panel with practical questions. For GPs, having a PN in the general practice who is largely responsible for performing the home visitation programme and who gained experience in it, was a positive development conducive to successful implementation.
Contamination control group (Q15)
None of the participating general practices in the control group had been involved in any form of proactive care for community-dwelling older people during the 18-month study period.
Interactional workability (Q16)
All PNs were satisfied with how they worked together with the GP in deciding about follow-up actions for detected problems. Yet one PN and one GP noticed that sometimes there was incongruence between the two of them: the proactive approach required GPs to address different kinds of problems and/or needs that otherwise might not have received attention at that point in time. Mostly, older people agreed with the suggestions done by the PN for follow-up actions. However, both PNs and GPs observed that quite a few older people were not willing to undertake any follow-up actions or only when problems had progressed substantially (incongruence).
Relational integration (Q17 + Q18)
At the end of the training programme, PNs felt very confident that they could perform a home visit independently (M = 6.27, SD = 0.91). In general, both PNs and GPs thought that the PN s knowledge, expertise or skills regarding care for older people were sufficient and, according to PNs, even increased during the intervention period. Those inexperienced with structural assessments would benefit from feedback, supplementary information, or examples on how to administer certain tests of the CGA and how to assign a score to the answers given by older people. A few PNs did not have that much expertise yet in deciding whether or not to undertake follow-up actions for detected problems and in formulating care and treatment plans correctly. One GP sensed a lack of affinity of the PN regarding care for older people. All older people believed the PN had sufficient knowledge about health, listened to them, took sufficient time for the home visit, and respected their needs and wishes.
Skill-set workability (Q19)
Both PNs and GPs considered the division of work regarding the home visitation programme clear and acceptable. As GPs often cannot attribute as much time to older people as they would like, the expertise gained by PNs in care for older people was very much appreciated. One PN was a little disappointed that the GPs did not use her expertise more often in arranging follow-up actions for older people.
Contextual integration (Q20 + Q21)
Six PNs felt that the home visitation programme was well integrated within the health care services offered by the general practice. According to two GPs, the home visitation programme fits within the health care services offered by general practices, as older people are familiar with the general practice. Furthermore, it enabled GPs to be the central care provider and to collect information that is relevant to them.
According to PNs and GPs, sufficient time for care for older people was an essential resource for adequate performance of the home visitation programme. PNs’ available time had to be carefully divided over various patient categories within the general practice. Five GPs thought that the time investment was disproportionate compared to the benefits in terms of detected problems. The other GPs thought the time investment was justifiable, as it yielded a lot more information about the older person.
Opinions of GPs diverged with respect to the importance of the reimbursement policy of health insurers for implementing care for (frail) older people. While some GPs considered it to be of minor importance, others believed its importance will grow over time due to competition between various disease management programmes for the available time of the PN. Finally, some GPs stated that continuation of the home visitation programme would largely depend on it.