Evaluation of the effectiveness of an ambulatory teaching/treatment programme for non-insulin dependent (type 2) diabetic patients

The aim of this study was to evaluate the effect of a structured teaching/treatment programme on the clinical and metabolic control of non-insulin-dependent (type 2) diabetic patients. The programme was aimed at improving the overall treatment quality in these patients through measures involving self-care, diet, exercise and weight reduction. Four theoretical-practical teaching units were given once a week to group of 5–8 ambulatory patients by previously trained general practitioners. Clinical and biochemical parameters were recorded at the beginning of the course and 1 year after its completion in 40 patients attending the programme and in 39 patients of similar clinical characteristics under conventional diabetes treatment, but receiving no structured teaching before or during the survey period (control group). The drop-out percentage in the intervention group (25%) was significantly lower than in the control group (45%),P<0.05, suggesting an incentive toward greater compliance in the former. At the end of the 1-year follow-up, the mean differences observed in the control and in the intervention groups were: body weight loss −2.4±0.5 kg vs −0.4±0.5 (P<0.001); haemoglobin HbA1 −0.2%±0.4% vs +0.8±0.4% (NS); number of daily oral hypoglycaemic agent intake −1.4±0.2 vs +0.9±0.2 tablets (P<0.001). Our results strongly suggest that this programme, applied through family doctors, may constitute an efficient tool to improve the compliance and clinico-metabolic control of type 2 patients at the primary health care level.


Introduction
Non-insulin-dependent (type 2) diabetes mellitus is a chronic disease mainly manifested in adults and frequently associated with other risk factors, particularly obesity [1, Correspondence to: J. J. Gagliardino

Patient selection and physician training
13 physicians invited to participate in a special 2-day seminar were instructed in the teaching/treatment programme. After the seminar, 9 out of these 13 decided to apply the structured programme to their own type 2 patients; another 9 physicians who had their medical practice in comparable health care units were invited to assist with the 1-year follow-up study as providers of control type 2 patients. This protocol was considered and approved by our hospital committee in accordance with the 1964 Helsinki Declaration. Accordingly, before participation in our study, all patients signed statements of informed consent. Previous experience had demonstrated the difficulties in selecting patients for the control and the intervention groups from the same physician/patient population. Our approach in patient selection was thus the following.
The first 6-7 type 2 patients consulting each physician were selected for inclusion in the intervention and in the control groups. In the latter, however, a larger number of patients was included based on the usual 50% drop-out observed in the type 2 patient population and in order to obtain a better match by age, gender and duration of diabetes with patients of the intervention group.
Patients included in both groups had comparable socio-economic levels and had previously received previous dietary advice from their physicians in a similar unstructured fashion and/or had been treated with oral hypoglycaemic agents (OHA) of the commercially available sulphonylurea type (glibenclamide). At the time of the study there were not biguanides on our market. Similar strict criteria were adopted in order to standardize between the two groups the frequency of visits to the physicians, the level of dietary caloric intake; and OHA prescription during the 1-year follow-up.
Exclusion-criteria for patient recruitment into this study were: (a) a newby diagnosed condition ofNIDDM; (b) age above 60 years; (c) presence of advanced microangiopathic complications, and (d) presence of other severe diseases (e.g. cancer).

Structure of the teaching and treatment programme
The course was given by the previously-trained physicians from the 2-day theoretical-practical seminar in a group teaching setting admitting up to 8 ambulatory patients. Every patient was encouraged to attend the course accompanied by spouse.
The programme consisted of 4 teaching units (90-120 min each) carried out once a week for 1 month.
During the first teaching unit, patients were introduced to the system of group discussion and asked to talk about their diabetes history; they were also instructed in the normal physiological range for serum glucose, the symptoms of hypoglycaemia and hyperglycaemia, and the renal threshold for glucose. As a practical part of this teaching unit, patients were trained to self-monitoring glycosuria with dry chemistry strips and to record the values.
The assay for glycosuria was chosen instead of the blood glucose test because we believe it is the easiest step to start the training of type 2 patients in their biochemical self-control and because the attainment of near normoglycaemia was not one of the aims of our study. The economy of test-strip cost was an additional consideration, not only for this study but also for the potential use in large populations at the primary health care level in the future.
At the end of this session, the patients were recommended to go on a very low caloric diet (600 calories) during the following week on alternating days until the following session, and to stop the intake of OHA in the meantime in order to prevent the potential risk of hypocaloric-induced hypoglycaemia. Such a regimen gives the patients the opportunity to test the effect of diet upon glucose levels. By contrast, the patients in the control group did not resort to such a low-calory diet. Those in the intervention group were also requested to monitor for glycosuria twice a day 2 h after the main meals and to record the findings in their log-books together with the daily measurement of their body weight.
During the second teaching unit, the patients were informed of and discussed the effect of obesity on peripheral insulin resistance and the advantages of a successful weight reduction. In order to reduce weight, patients learnt to divide food into three categories: very helpful, convenient and inconvenient. An individual meal planpractically self-designed by every patient -based on the ingestion of about 1000 Cal/day was discussed. Patients were recommended to adopt this meal plan as a long-run daily objective until attaining the desiderable body weight. M. I. Domenech et al.: Ambulatory teaching/treatment programme At the beginning of the last two teaching sessions, the patients were requested to explain and discuss with the other participants their experiences after undertaking the changes in diet, body weight and self-monitoring glycosuria which occurred during the previous weeks.
During the third teaching unit, the main topics were foot care and physical activity. The physician -or educator -also performed patient feet examinations.
At the fourth session, basic rules to be applied on "sick days" (intercurrent episodes of an acute disease) were explained, as well as the minimal clinical and biochemical tests required for effective control and follow-up of the patient.
The educational material consisted of: (a) a set of 25 coloured flip-charts, (b) teaching files for the educator as a structured guideline on how to carry out each session, (c) a set of 50 photographs of different foodstuffs representing 100 calories each, (d) question cards for distribution among class attendants as a standardized procedure to verify the knowledge acquired in previous sessions, (e) an individual log-book for recording all the self-monitored data (glycosuria, body weight), (f) a patient booklet including the main contents and other aspects of the programme, and (g) a questionnaire for evaluation and documentation of the patient diabetes-related knowledge before and after the programme.
This set of material was kindly provided by Boehringer Mannheim to each physician after their training seminar.

Evaluation protocol
The following parameters were examined at the beginning of and 1 year after the course: diabetes-related knowledge as revealed by a 20-question questionnaire; changes in body weight, number of OHA and glycosylated haemoglobin (HbA1) levels employed daily, and as measured with the HbA I test combination kit of Boehringer Mannheim. The values of the latter test allowed us to gain a rough idea of the credibility of the glycosuria values recorded by the patients during the follow-up period and the degree of metabolic control achieved.
The same follow-up protocol (except for the questionnaire on diabetes related knowledge) was used in the control group. In this regard, a simple comparison of patient knowledge within the intervention group before and after the programme provided a sufficient control for this parameter.

Statistical analysis
The intervention and control groups were compared initially and after 1 year; the chi-square test (with Mantel-Haenszel P value), the Pearson's correlation coefficient and the Wilcoxon method paired signed ranks test were used for statistical data analysis. Results were regarded as significant if P < 0.05.

Group characterization
Mean age and duration of diabetes, percentage of men and women as well as other clinical and biochemical parameters were similar in both groups (Table 1); although the mean body mass index (BMI) was somewhat lower in the control than in the intervention group, the difference was not statistically significant.
The intervention group initially included 53 patients; 40 of them (75%) were re-examined after 1 year. Dropouts were a result of the following factors; one death from an intervening pulmonary disease; one psychological de-pression; one change of permanent residence; one family problem; six refusals to participate in the follow-up examination; and three failures to complete the four units of the programme.
In the control group, 39 out of 71 patients (55%) were re-examined at the end of one year. Such a drop-out percentage, being significantly higher than that occurring in the intervention group (Z 2, P < 0.05), was within the expected range for our outpatient system. We were unable to identify the causes for these particular individual drop- From these data we calculated the percentage of patients in each of the two groups who showed an improvement (i.e., a decrease) in the above parameters at the end of one year. For this purpose and based on the coefficient variation of our measurements, we recorded only a body weight loss greater than 0.5 kg and a decrease in HbA 1 of  more than 0.5% (Fig. 1). Accordingly, within the intervention group, 80% of the patients (Fig 1) were found to exhibit a significant body weight loss (mean vlaue -3.7_+ 0.4 kg, P<0.05). We also noted a significant correspondence between this parameter and a concomitant reduction in daily OHA intake (Pearson's correlation coefficient, P <0.002). The glycosylated-haemoglobin levels of these individuals moreover remained within an acceptable range according to the criteria of the European NIDDM Policy Group: (i.e., 9.1 +0.4%) [6]. In contrast to this pattern, only 46% of the subjects of the control group showed a reduction in body weight (Fig. 1), with their mean decrease being -3.1_+0.5 kg (P< 0.05). The difference between this value and the corresponding percentage for the intervention group was highly significant (Z 2, P<0.002). Furthermore, the weight loss within the control group was

Data obtained before the beginning of the instruction programme vs after one-year follow-up for selected subpopulations within the intervention and control groups.
A Values for HbA 1 levels, body weight, and daily OHA intake as measured within the specific subgroup of patients who evinced a decrease in the first of these parameters (data in boldface); B values for daily OHA intake, body weight, and HbA 1 levels as measured within the specific subgroup of patients who evinced a decrease in the first of these parameters (data in boldface) not accompanied by any overall change in the subjects' HbA 1 levels or a concomitant dimininution of their daily OHA intake. We next decided to treat each of the parameters other than body weight -HbA 1 levels and OHA intake -in turn as an independent variable so as to examine any correlations that might exist between it and either body weight or the remaining third factor. For this purpose, we thus pooled the data from only those patients who exhibited a significant decrease in the parameter being isolated after the one-year follow-up period and then calculated the mean values for the other two factors within that subgroup for comparison (Table 2).
Within the intervention group, 43% (Fig. 1) evinced an actual decrease in the HbA 1 values (mean difference, -2.7%; P<0.005; Table 2A). This metabolic improvement was, in turn, accompanied by a significant body weight loss and a reduction in the daily OHA intake (Table 2A). By contrast, about one-third of the patients within the control group (Fig. 1) exhibited a significant decrease in HbA1 values (mean difference, -1.4%; Table 2A), though in their circumstance this decrement was accompanied by only a marginally significant weight loss (mean difference, -1.0 kg) and a slight increase in the daily intake of OHA (mean value, +0.3 tablets; Table 2A). It should be noted that, although the percentage values for the individuals showing diminished HbA 1 levels at the end of one year were statistically comparable between the two groups (i.e., 43% vs 33%, Fig. 1), the average decrease in the parameter itself at that time (Table 2A) was, in fact, significantly greater (Z2, P<0.05) in the intervention group (25%) than in the controls (14%).
Of the subgroup of patients taking OHA at the beginning of the study, 23 out of the 29 within the intervention group (79%, Fig. 1) decreased their daily intake at the end of one year. These individuals also exhibited a slight reduction in their HbA 1 levels, and their average weight loss was comparable to that of the intervention group as a whole (-3.8 kg). Of the 34 control subjects undergoing OHA treatment, however, only 6 (18%, Fig. 1) managed to diminish their daily intake of the drug (Z2), with respect to the intervention group, P<0.0001). The magnitude of this decrease in OHA consumption was also significantly smaller in these patients than in those of the intervention group (-1.5 vs -0.9 tablets per day, respectively; P<0.05; Table 2B). Moreover, although they too, as a whole, diminished in body weight, the average extent of this loss was less than one-half than that of the intervention-group subjects (-1.7 vs -3.8 kg, Table 2B). Finally, by the end of the one-year follow-up, the fraction of the patients not taking OHA rose from 28% to 58% in the intervention group, as opposed to the modest increase of from 13% to 21% seen for the control subjects (P value for the comparison of the difference in these percentages between the two groups <0.05).

Discussion
The fact that in our study 80% of the patients in the intervention group still had a significant body weight decrease M. I. Domenech et al.: Ambulatory teaching/treatment programme at the end of the 1-year follow-up shows that the implementation of a structured group education programme decreases the usual treatment drop-out and widens the traditional compliance with the nutritional plan. These results are comparable to those obtained by Kronsbein et al. in a German population with a similar programme [7].
It is by now well established that a loss in body weight decreases the insulin-resistant state of a NIDDM patient, thus producing an improvement in the associated metabolic imbalance [2,[8][9][10][11][12][13]. Similar changes were observed in our intervention group, where improvement in HbA 1 values was accompanied by a simultaneous decrease in body weight and daily OHA intake. Moreover, 58% of the patients in this group was no longer on OHA treatment 1 year after the introduction of a reducing diet. This percentage of patients that could control their disease only through a suitable diet treatment was significantly higher than that found in the control group (21%) and even greater than the 10%-30% reported in the literature [14].
The decrease or suppression of the daily OHA intake in type 2 diabetes is very important because: (a) B cells may undergo a process of desensitization to these secretagogic agents which, in the course of time, decrease their therapeutic effect [15]; (b) the stimulation of insulin by OHA contributes to the functional overburden that B cells already have and consequently to their exhaustion, and (c) the decrease (or suppression) of OHA contributes to lowering the costs in the treatment of the disease and nullifies the potential risk of suffering hypoglycaemic episodes [ 16].
In the intervention group, the percentage of patients who showed a decrease in HbA 1 was larger but not significantly different from that observed in the control group. However, the mean decrease in HbA 1 in this intervention group was significantly larger than that in the corresponding controls. This would suggest that if we want to improve the current percentage of successful treatment, we should increase some of the programme requirements, such as a greater weight loss, a more precise adjustment of OHA dosage and a regular self-monitoring of blood glucose levels.
Our results also show that this programme, applied through family doctors, constitutes an efficient tool for improving the clinical-metabolic control in type 2 patients at the primary health care level. This type of programme can be developed not only by family doctors but also by other members of the health care team such as dietitians and previously-trained nurses, and implemented on a large scale as recently shown by Gruesser et al. [17]. Such an implementation may contribute to preventing the development and progression of the chronic complications in NIDDM, thus decreasing the high direct [18] and indirect [19] costs of the management of this disease.