By Martin Crump.
Martin is a Director and co-founder of Evolution. He is a certified NLP Master Trainer with a wealth of experience of working with organisations of all sizes and types across the UK.
I was approached in 2007 by Dr Jeorg Beyer who thought that NLP training of Diabetes Clinicians may have an effect on the health outcomes of their patients.
It was a great opportunity to measure any benefits of NLP influencing techniques as there is a very measureable evaluation of change through blood test.
I trained a number of Diabetes professionals – GPs and Diabetes nurses who applied their NLP skills and knowledge with their patients.
Dr Beyer checked blood test results for a year before the training and compared them to the year following the training with 94 patients.
He concluded that there was a statistically significant chance that the NLP training led to clinical improvements.
His report is below:
Does NLP improve outcomes in patients with Type 2 Diabetes Mellitus – A case-crossover study
Dr CJ Beyer
Dovecot Health Centre, L14 0NL, United Kingdom
Background:
Patient concordance with doctors’ recommendation is poor. Diabetes education has been accepted in diabetes care but the effect of diabetes education on glycaemic control and the components of education responsible for such an effect are uncertain.
The Oxford English Dictionary defines Neuro-linguistic programming (NLP) as “a model of interpersonal communication chiefly concerned with the relationship between successful patterns of behaviour and the subjective experiences (esp. Patterns of thought) underlying them” and “a system of alternative therapy based on this which seeks to educate people in self-awareness and effective communication, and to change their patterns of mental and emotional behaviour”.
UKPDS 35 has shown that each reduction in HbA1c by one percent point reduces the risk of death related to diabetes by 21%.
We hypothesised that using NLP during consultations will make clinicians more effective at convincing patients to make lifestyle changes and adhere to treatment which should be reflected in changes of the glycosylated haemoglobin (HbA1c).
Objective: To evaluate whether a two day course in NLP for health care professional improves outcomes in patients with Type 2 Diabetes Mellitus
Design: Case-crossover study
Setting: General Practice in a deprived inner city area in the United Kingdom
Patients: 94 patients with Type 2 Diabetes Mellitus
Measurements:
Comparison of the average change of the HbA1c within the year before and the year after a course in NLP lasting two full days one month apart.
Results:
The mean difference of the HbA1c within twelve months before NLP was -0.006 (95% confidence interval -0.068 to 0.081), the mean difference of HbA1c within twelve months after NLP was -0.28 (95% confidence interval -0.35 to -0.21, N=93).
Conclusions:
NLP may lead to a clinically and statistically significant improvement of the HbA1c. It is likely that the improvement is due to NLP because all patients have been reviewed by the same practice nurse before and after she attended the NLP course and no changes in treatment algorithms have been made, although it is very difficult to know whether our practice nurse has always used NLP during her consultations. As the patients have been their own controls, the study design controls for characteristics of patients that may affect concordance with treatment which do not change over a period of time.
Abbreviations: NLP – Neuro Linguistic Programming, HbA1c – glycosylated Haemoglobin
Background:
It is well known that patient concordance with doctors’ recommendation is poor. Diabetes education has largely been accepted in diabetes care but the effect of diabetes education on glycaemic control and the components of education responsible for such and effect are uncertain.
The Oxford English Dictionary defines Neuro-linguistic programming (NLP) as “a model of interpersonal communication chiefly concerned with the relationship between successful patterns of behaviour and the subjective experiences (esp. Patterns of thought) underlying them” and “a system of alternative therapy based on this which seeks to educate people in self-awareness and effective communication, and to change their patterns of mental and emotional behaviour”.[1]
UKPDS 35 has shown that each reduction in HbA1c by one percent point reduces the risk of death related to diabetes by 21%.[2]
We hypothesised that using NLP during consultations will make clinicians more effective at convincing patients to make lifestyle changes and adhere to treatment which should be reflected in changes of the glycosylated haemoglobin (HbA1c).
Our objective was to evaluate whether a two day course in NLP improves outcomes in patients with Type 2 Diabetes Mellitus.
Setting:
This case-crossover study has been performed in a two partner practice in Liverpool in an area with a generally more deprived and poorer health status. The practice provides General Medical Services for approximately 3800 patients. Population age breakdowns and birth rates are similar to the Liverpool average. The main causes of death are ischaemic heart disease and cancers.
Methods:
The author conducted a case-crossover study. The design of case-crossover studies has been shown to apply best if the exposure is intermittent (disease reviews by health care professionals trained in NLP), the effect on risk is immediate and transient, and the outcome is abrupt (intermittent measurements of HbA1c as a proxy for risk reduction).[3]
Both General Practitioners and the Practice Nurse performing the annual diabetes reviews and follow-up of most patients with type 2 diabetes mellitus participated in a course in NLP on two full days one month apart (October and November 2007). The instructor is a Master Trainer in NLP.
The author compared the mean difference between the last HbA1c value and the first HbA1c value within a year before the NLP course (HbA1c2-HbA1c1 (Illustration 1)) and the mean difference of the last value within one year after and the last value before the NLP course on the same patient population (HbA1c3-HbA1c2 (Illustration 1)). As the patients have been their own controls, the study design controls for characteristics of patients that may affect concordance with treatment which do not change over a period of time.
Additionally, age and gender distribution and number of patients with at least two HbA1c recordings in the year before the NLP training and at least one recording in the following year as well the mean, 95% confidence intervals, median and mode of the first HbA1c one year before and the last just before NLP have been calculated.
Results:
Table 1 shows mean, median, mode and 95% confidence intervals of the HbA1c one year before NLP as well as the number of patients (N) with complete (93), incomplete (46) or without (27) recordings of HbA1c before and after NLP.
| HbA1c one year before NLP | Mean | Median | Mode | 95% Confidence interval | N |
| With complete recordings of HbA1c before and after NLP | 7.0 | 3.8 | 6.9 | 6.6 – 7.6 | 93 |
| No recording of HbA1c after NLP | 6.5 | 6.3 | 6.3 | 5.9 – 7.2 | 46 |
| No recording of HbA1c before NLP | – | – | – | – | 27 |
| Total | 6.9 | 6.6 | 6.3 | 6.5 – 7.3 | 139 excluding 27 patients without recordings of their HbA1c one year before NLP |
Table 1: HbA1c one year before NLP
Of the 166 patients on our Diabetes Register, 93 had at least two HbA1c recordings in the year before the NLP training and at least one recording in the following year. The average HbA1c in patients with complete recording s before and after NLP has been slightly higher than the average HbA1c were incomplete. Although it is conceivable that patients, whose diabetes is well controlled are not followed up as often as those whose diabetes is not well controlled, 95% confidence intervals show that this difference is not statistically significant.
Forty-six patients were male and 47 patients female aged 40 to 89 years (mean = 67 years, median = 70 years, mode = 74 years).
The mean difference of HbA1c within twelve months before NLP was -0.006 (95% confidence interval -0.068 to 0.081), the mean difference of HbA1c within twelve months after NLP was -0.28 (95% confidence interval -0.35 to -0.21, N=93)
Discussion:
Diabetes education has largely been accepted in diabetes care but the effect of diabetes education on glycaemic control and the components of education responsible for such an effect are uncertain. A meta-regression analysis on 28 educational interventions (n=2439) revealed that face-to-face delivery, cognitive reframing teaching method and exercise content collectively explained 44% of the variance in glycaemic control – the net glycaemic change being 0.32% lower in the intervention group than in the control group.[4] This is in keeping with the findings with our patients.
Norris et al. observed in their meta-analysis of 31 trials that self-management education on adults with type 2 diabetes improves glycosylated haemoglobin levels at immediate follow-up by 0.76% (95% CI 0.34 to 1.18) more than the control group, declines by 0.26% (95% CI – 0.73% decrease to 0.21% increase) one to three months after intervention and by 0.26% (95% CI 0.05% to 0.48%) at more than three months. The authors conclude that learned behaviours change over time.[5]
Mean and 95% confidence intervals of the HbA1c one year before NLP of patients with and without complete recordings of HbA1c support the assumption that patients with well controlled diabetes are followed up less often than patients with poorly controlled disease. Selection bias however is unlikely as the 95% confidence intervals of the two groups of patients overlap.
Similar to Norris et al. and Ellis et al. the author observed a mean change of -0.21 to -0.35 percent points. This change is certainly clinically significant as it would lead to a reduction in risk for deaths related to diabetes of four to seven percent. The 95% confidence intervals of the mean difference of the HbA1c before and after NLP suggest that the improvement after NLP is also statistically significant.
Cost analysis indicates that where costs associated with patient education were in the region of £500-600 (€570 -683, US$813 – 975) per patient, benefits over time would have to be very modest to offer an attractive cost-effectiveness profile. [6]
The cost for the NLP trainer for fifteen delegates for 2 full days was around £2000 (€2270, US$3245) and is therefore likely to be cost effective.
NLP may lead to a clinically and statistically significant improvement of the HbA1c. It is likely that the improvement is due to NLP because all patients have been reviewed by the same practice nurse before and after she attended the NLP course and no changes in treatment algorithm have been made, although it is very difficult to know whether our practice nurse has always used NLP during her consultations.
To learn more about how NLP can help you get better results in your work visit the NLP At work confernce in September 2011
[1] Oxford English Dictionary, Oxford University Press 2003
[2] Irene M Stratton, Amanda I Adler, H Andrew W Neil, David R Matthews, Susan E Manley, Carole A Cull, David Hadden, Robery C Turner, Rurz R Holman on behalf of the UK Prospective Diabetes Study Group. BMJ 2000;321:405-12
[3] M. Maclure and M. A. Mittleman. Annual Review of Public Health May 2000;21:193-221
[4] Ellis SE, Speroff T, Dittus RS, Brown A, Pichert JW, Elasy TA, Patient Educ Couns. 2004 Jan;52(1):97-105
[5] Norris SL, Lau J, Smith SJ, Schmid CH, Engelgau MM. Diabetes Care 2002;25(7):1159-6
[6] Loveman E, Cave C, Green C Royle P, Dunn N, Waugh N. Health Technol Assess. 2003;7(22)iii:1-190
