Within trial analyses have shown that there were no significant overall cost differences between nurses and lay-trainers. However there were significant differences between the two groups with regards to the cost of training (significantly higher for lay educators) and the cost of delivering the trial consultations (significantly lower for lay educators). There was a cancelling out effect of the two components of the intervention with lower costs for nurses in the training part and lower costs for lay educators in delivering the consultations, however there was no overall difference in effect (frequency of healthcare utilisation events). Although the asthma trainers were more expensive to train per patient the trial results show that they can deliver asthma consultations at a much lower cost. The provision of robust training is essential to the safety, acceptability to practices and confidence of the lay educators. The lay educators were providing direct care and substituting for a practice nurse.
In this analysis we have not included the cost of basic nurse training. We set out to assess the additional extra cost of skilling either a lay person or a nurse to competently undertake an asthma review. Hidden costs such as the cost of mentoring and supporting the practice nurses and lay educators in practices were not explored, although it was noted that extensive support was given at the beginning of the trial. To extrapolate this study to real life the lay educators would need some form of mentoring within their practices. Due to the under-recruitment in the Manchester arm of the study the results, whilst showing no differences in outcomes between the two arms, did not actually have sufficient power to demonstrate equivalence.
The cost of follow-up telephone consultations incorporated staff time, but did not include the administrative cost of calls. These costs would have been similar in both nurses and lay educators. Since this study has been carried out there has been an increase in tele-health and perhaps using texts or reminders may have increased the numbers of telephone consultations carried out. The increase in mobile phones usage in comparison to landlines would also be interesting to investigate further.
As discussed in the original trial using lay educators to deliver health care, the study had a number of logistical problems. Several of the trainers dropped out of the study shortly after training, significantly contributing to the training costs. This was in part due to a delay in the trial start-up. The asthma trainers were not contractually employed and the training course was a “one-off” making it difficult to re-recruit new trainers. Updating of training and training of new staff should be considered if lay educators are rolled out into clinical practice. Prior to the trial, and contrary to expectations, very few of the practice nurses had had any additional asthma training (only one of the forty-six nurses had an Asthma diploma). This low rate of acquisition of training to permit nurses to competently undertake additional tasks such as asthma reviews may or may not be representative of other parts of the UK . Due to competing pressures for time many practice nurses found it difficult to conform to the trial protocol, as is seen with the lower completed consultation rates compared with the lay educators. The lay educators for example, were more successful than nurses in contacting patients and persuading them to attend for review and undertook more telephone follow ups. In reality it is likely that the nurses did not adhere to the recommended time durations for consultations which would have reduced their costs as we have considered in our scenario analysis, although it is likely that lay educators might also “speed up” with time. Certainly unit costs for the lay educators could be expected to be reduced in real life, since within the trial, the lay educators often travelled to a practice to see only one or two patients randomised to them, whereas a lay educator working full time and in one location and preferably across practices could be more effective.
These scenarios were devised assuming that changes in intervention delivery would not impact on unscheduled health care utilisation. Scenario 1 is the most likely scenario to have an impact of the unscheduled healthcare utilisation. However full protocol adherence was in both arms of the study and improvement in outcomes was similar and therefore no significant difference is seen.
This trial, if completed according to protocol would have represented one of the largest trials of review and self management support ever undertaken. The difficulties with recruitment in Manchester, accompanied by problems with Ethics Committee approvals and delays in utilising lay educators newly acquired skills, led to a situation where there was drop out of educators, practices and nurses and full equivalence of intervention was intimated if not quite proven. However few seeing these results can doubt that in one prescribed area of competency, well trained lay educators have been shown to be capable of delivering healthcare in a manner comparable to traditional care, which in the UK is by primary care based Practice Nurses. A lay-led asthma self management education package may not be possible in all healthcare systems, particularly where patients expect to receive care from a medically qualified individual or where there is a high level of medical dominance.
Practicalities of implementing lay educators into clinical practice
This study has shown that there was no difference in the cost of unscheduled healthcare between the nurses and the lay educators. It would not be possible to replace nurses completely within general practices as only a small proportion of their time is spent on patients with asthma, one study has shown that this is as little as 7% of their time . Potentially we could free up clinical time to see other patients and perform other duties but it would not be in effect a cost saving because of additional spending on lay services while nurses are still employed.
It should be considered that using lay educators would incur extra costs providing an additional service albeit at a significantly lower cost than nurses. Other factors should also be considered, we found in the original study that there was a significant number of lay educators who dropped out before and after training and high levels of practice nurse turnover. Lay educators would have to appropriately recruited and trained to ensure training costs are not wasted. Training would need to be reinforced and perhaps repeated annually to ensure adequate skill levels which would add to the running costs. Ultimately lay educators would need to be reliable and as members of the public may be more of an unknown quantity than already employed nurses.
Lay educators will also need to be employed on a formal contract and would need continued mentoring and guided support to avoid isolation, both of these issues were raised during the trial. It is crucial that the role of the lay educator is well defined and the employment contract is well defined as shown by Brown et al.. In this study lay educators were responsible for claiming back the costs for their time at £8 per hour. The employment contract was not well defined and it is difficult to perceive whether individuals were acting as an expert member of staff or volunteer. This has implications on the success of the individual at delivering the outcomes and more research needs to be carried out on the role, expectations and effectiveness of expert patients in trials and clinical practice as well as the roll-out cost implications. It would be considerably cheaper to use lay educators as a volunteers but further work on the implementation of lay educators into “real-world” clinical practice needs to be carried out to ensure this is the best option.