Int J Clin Pharm DOI 10.1007/s11096-015-0122-3
RESEARCH ARTICLE
Pharmaceutical care of adolescents with diabetes mellitus type 1: the DIADEMA study, a randomized controlled trial Emina Obarcanin1 • Manfred Kru¨ger2 • Petra Mu¨ller3 • Verena Nemitz3 • Holger Schwender4 • Snijezana Hasanbegovic5 • Sena Kalajdzisalihovic5 • Stephanie La¨er1
Received: 22 June 2014 / Accepted: 14 April 2015 Koninklijke Nederlandse Maatschappij ter bevordering der Pharmacie 2015
Abstract Background Physiological and psychological changes during puberty and a low adherence to complex treatment regimens often result in poor glycemic control in adolescents with type 1 diabetes mellitus (T1DM). The benefit of pharmaceutical care in adults with diabetes mellitus type 2 has been explored; however, evidence in adolescents with T1DM is scarce. Objective To evaluate the impact of pharmaceutical care in adolescents with T1DM provided by pharmacists, in collaboration with physicians and diabetes educators on important clinical outcomes (e.g., HbA1c and severe hypoglycemia) Setting: At the outpatient Helios Paediatric Clinic and at the 12 regular community pharmacies of the study patients with 14 pharmacists in the Krefeld area, Germany, and at the University Pediatric Clinic with one clinical pharmacist on-site in Sarajevo, Bosnia-Herzegovina. Methods A randomized, controlled, prospective, multicenter study in 68 adolescents with T1DM. The intervention group received monthly structured pharmaceutical care visits delivered by pharmacists plus supplementary visits and phone calls on an as needed basis, for 6 months. The control group received usual diabetic care. Data were collected at baseline and after 3 and 6 months. Main outcome measures: The
between-group difference in the change from baseline in glycosylated hemoglobin (HbA1c) and the number of severe hypoglycemic events in both groups. Results The improvement from baseline in HbA1c was significantly greater in the intervention group than in the control group after 6 months (change from baseline -0.54 vs. ?0.32 %, p = 0.0075), even after adjustment for country-specific variables (p = 0.0078). However, the effect was more pronounced after only 3 months (-1.09 vs. ?0.23 %, p = 0.00002). There was no significant between-group difference in the number of severe hypoglycemia events. (p = 0.1276). Conclusion This study suggests that multidisciplinary PhC may add value in the management of T1DM in adolescents with inadequate glycemic control. However, the optimal methods on how to achieve sustained, long-term improvements in this challenging population require further study. Keywords Adolescents Bosnia-Herzegovina Germany Glycemic control Pharmaceutical care Pharmacists Type 1 diabetes mellitus
Impacts on Practice & Emina Obarcanin [email protected] 1
Department of Clinical Pharmacy and Therapeutics, Heinrich Heine University, Universita¨tsstrasse 1, 40225 Du¨sseldorf, Germany
2
Linner Pharmacy, Krefeld, Germany
3
Helios Pediatric Clinic, Krefeld, Germany
4
Mathematical Institute, Heinrich Heine University, Du¨sseldorf, Germany
5
Pediatric Clinic, University Clinical Center Sarajevo, Sarajevo, Bosnia-Herzegovina
•
•
•
Pharmaceutical care may contribute to the improvement of glycemic control in adolescents with type 1 diabetes mellitus. Implementation of pharmaceutical care in a multidisciplinary team may help to prevent acute and, potentially, long-term complications of T1DM. The pharmaceutical care protocol developed for adolescents with T1DM may help to bridge the gap in healthcare provision across different healthcare systems, especially in countries where healthcare resources are limited.
123
Int J Clin Pharm
Introduction Type 1 diabetes mellitus (T1DM) incidence in children and adolescents is increasing worldwide [1–4]. Physiological changes with greater insulin resistance during puberty, as well as psychological factors and non-adherence to insulin regimens, result in poor glycemic control [5–8], leading to increased morbidity, and premature mortality [5, 9]. T1DM complications represent a major economic burden [10]. A multidisciplinary approach with intensive patient monitoring is required to prevent complications of diabetes and improve quality of life [11]. Multidisciplinary pharmaceutical care (PhC) programs have improved outcomes in adults with type 2 diabetes mellitus (T2DM) [12–14]. However, data on such programs in adolescents with T1DM remain scarce. A randomized, controlled, French study with pharmacists faxing glucometer data to the hospital revealed no significant difference in HbA1c between the study groups [15]; an observational American study in which pharmacists directed a peer-support program showed even an increase in HbA1c [16]. These findings indicate that pharmacist interventions alone, in the absence of PhC (as defined in Standards of Practice [17]) are not sufficient to improve outcomes in this population. Therefore, further study was needed to evaluate the effects of integrating a structured PhC program [17] in adolescents with T1DM. Our DIADEMA (DIabetes in ADolescence Engagement and Monitoring by phArmacists) study, designed according to Standards of Practice [17], incorporated clear responsibility assignment, counseling, and outcome evaluation [18]. This involved: (1) assessment of drug-related needs through data collection and interpretation, (2) development of Pharmaceutical Care Plans (PCPs) with problem-solving interventions and individual goals, (3) follow-up and outcome evaluation. We hypothesized that implementing this multidisciplinary, structured PhC would improve HbA1c and other clinical outcomes.
Aim of the study We evaluated the impact of intense and structured diabetesrelated pharmaceutical care on clinical outcomes in adolescents with T1DM in Germany and Bosnia-Herzegovina. Ethical approval The study was approved by the Ethics Committee of the Medical Faculty of Heinrich-Heine University, Du¨sseldorf, Germany (No. 3991). The Agency for Medicinal Products and Medical Devices of Bosnia-Herzegovina and the Ethical Committee of the Clinical Center at the University
123
of Sarajevo did not require ethical approval, as they did not qualify this study as a clinical trial of drugs (Reference No. 10-07-56-1207/12), (Official Gazette Nr. 4/10) [19, 20]. The study was performed according to the Declaration of Helsinki and Good Clinical Practice (GCP) guidelines. Study patients were verbally informed about the study and patients and their parents signed written consent prior to the start of the study.
Methods Study design This was a randomized, controlled, prospective 6-month trial conducted in adolescents with T1DM and poor glycemic control. Participants and setting Eligible patients (12–18 years of age) who had confirmed T1DM diagnosed C6 months prior to enrolment were outside the partial remission phase, as indicated by insulin use of C0.5 IU/kg/day [21], used intensified conventional therapy (ICT) or continuous subcutaneous insulin infusion (CSII) and had C2 measurements of HbA1c consistently C7.5 %, taken C3 months apart. Exclusion criteria included intellectual and developmental disabilities, drug abuse, psychiatric conditions, pregnancy, and breastfeeding. The study had two sites: Pediatric Clinic Helios Hospital Krefeld in Germany (with 12 community pharmacies and 14 pharmacists) and University Pediatric Clinic Sarajevo in Bosnia-Herzegovina (with one on-site clinical pharmacist). Recruitment and randomization The clinics recruited eligible patients. Patients in Germany named their regular community pharmacies, from which 14 pharmacists were recruited. A post hoc power analysis based on the actual sample and effect size was performed using GPower version 3.1 [22]. Randomization was performed by diabetes educators (nurses) using a simple, easy and well analyzed coin-toss method [23–25]. Physicians supervised the coin toss to avoid potential bias [24, 25]. Nurses wrote eligible patient names and sealed these in envelopes to conceal allocation and minimize selection bias [24, 26]. The coin toss resulted in 40 ‘‘tails’’ (predesignated as the intervention group) and 29 ‘‘heads’’ (control). Following randomization, envelopes in each group were consecutively numbered by nurses and opened. To improve willingness to participate, patients who
Int J Clin Pharm
consented verbally were randomized and then signed consent prior to the study start.
study coordinators (M. Kru¨ger and E. Obarcanin) on PhC, T1DM management and GCP prior to and during the study.
Data collection
Control group
Diabetes educators (nurses) and pharmacists collected and exchanged data using two-part CRFs. Participants attended the clinics 3 times (baseline and 3 and 6 months) to receive usual care (usually every 3 months), including HbA1c measurement and counseling by the diabetes team. Nurses recorded socio-demographic and other data (e.g., HbA1C, insulin dose) at each visit, using ‘‘Physician CRFs’’; these were verified by physicians, according to GCP. The CRFs of the intervention group were accessible on-site in Bosnia and were faxed to the pharmacists in Germany.
Participants had scheduled diabetic clinic appointments every 3 months. As part of the ‘‘usual care visits’’ in the clinics, patients were advised to measure their BG at least 4 times daily, [21] and the frequency of BG monitoring was routinely documented by diabetes educators (nurses). During the study, patients could request counseling from their local pharmacists in special cases (e.g. school excursions, hypoglycemia). However, as these local pharmacists were not study participants, we could not confirm if counseling was requested or received.
Intervention group
Primary outcome
Patients received monthly (over 6 months) scheduled PhC visits (60-90 min) with pharmacists, plus on-demand phone calls and additional visits on an as needed basis, as required by patients and in special situations (e.g., repeated uncontrolled hyperglycemia, school excursions, frequent hypoglycemia). The PhC was standardized in both countries according to the Standards of Practice [18]. During the visits pharmacists recorded clinical data (i.e., blood glucose (BG), hypoglycemia, ketone values, insulin therapy and dose, tolerability) using ‘‘Pharmacist CRFs’’. Patients in the intervention group measured and documented their BG at least 4 times daily (i.e., fasting, after administering insulin, bedtime) [21], preferably 5–6 times to include measurements 2 h after meals or during or after sports. Based on data and evaluation of BG records [21], pharmacists assessed drug-related needs and identified problems (e.g., insulin non-adherence at school, dose adjustment for hyperglycemia or sports). Individual Pharmaceutical Care Plans (PCPs) were developed for each patient, including problem-solving interventions, (e.g. dose adjustments, diet modification, contacting teachers) and C1 measurable goal (e.g., HbA1c reduction in 3 months). Immediate insulin dose changes pharmacists proposed to and discussed with physicians personally (in Bosnia) or by phone (in Germany). Every PCP of visits 3 and 6 had to be faxed/submitted and discussed with physicians. In Germany, multidisciplinary face-to face meetings between the community pharmacists, physician and diabetes educator took place before and at the study start and every 2 months during the study. Pharmacists also provided education on standardized T1DM-related topics: treatment goals, insulin use, self-monitoring of blood glucose (SMBG), prevention of complications, and physical activity [27]. Follow-up and progress evaluation was performed monthly. Individual and group training were provided to pharmacists by the
The primary endpoint was the change in HbA1c from baseline at 3 and 6 months. Secondary outcomes (both groups) Mean number of severe hypoglycemic events, defined as low glucose requiring assistance of another person [28], was recorded by nurses from multiple sources (e.g., medical records, hospitalizations, BG records, and glucometer data) and was retrospectively analyzed for the 6 months prior to the study, and during the study. Well-being Index (WHO-5) (intervention group) A validated instrument was used to assess psychological status [29, 30] at months 3 and 6. The five positively worded items (i.e., ‘‘I have felt: cheerful spirits; calm; active; I woke up fresh; my day is interesting’’) measure well-being over the past 2 weeks using a 6-point Likert scale from 0 (not present) to 5 (constantly present).Total scores were multiplied by 4 to obtain percentage scores of 0–100, with higher scores indicating better well-being. Scores \50 and \28 % indicate low mood and depression, respectively. Repeated scores of \28 % were conveyed to the patient’s physician, with a clinical psychologist consult if necessary. Data analysis Analysis of data was performed in the per-protocol population (Fig. 1). Statistical analysis, including baseline characteristics and HbA1c data collected at baseline and after 3 and 6 months, was performed using the statistical software environment R, version 3.0.1 [31]. The Mann– Whitney test for quantitative (non-normally distributed)
123
Int J Clin Pharm
Screening
Assessed for eligibility (n = 80)
Not included (n =11) Not meeting inclusion criteria (n =3) Declined to participate (n =8)
Allocation
Randomization
Fig. 1 Patient flowchart of the DIADEMA trial
Randomized (n = 69)
Allocated to pharmaceutical care (n = 40) Bosnia (n = 26); Germany (n = 14)
Analysis
Follow-up
Received allocated intervention (n = 40)
Discontinued intervention (n =1) Declined to participate after 3 months due to lack of motivation
Pharmaceutical care group Analyzed (n =39) Excluded from analysis (study withdrawal) (n =1)
variables was used to analyze primary and secondary outcomes, and the v2 or Fisher’s exact tests were used to analyze categorical variables in the secondary outcomes. The p-values were calculated and used to interpret the statistical significance of the differences between the groups. For the primary outcome analysis, the Mann–Whitney test was used, as the differences in HbA1c values at different time points were non-normally distributed or only marginally normally distributed. Furthermore, a linear regression analysis and analysis of variance with the Wald test were applied to test whether the variables ‘‘country (Bosnia/Germany) and ‘‘Group’’ (intervention/control) significantly influenced HbA1c changes. Further analysis of the well-being (WHO-5), included linear models to test if baseline covariates such as age, diabetes duration, sex, and insulin device influenced the results.
123
Allocated to usual care (n =29) Bosnia (n=20); Germany (n =9) Excluded, not meeting inclusion criteria (n =1) Finally allocated to usual care (n =28)
Lost to follow up (n =2) Did not come to the control visit after 6 months
Usual care group Analyzed (n =26) Excluded from analysis (missing outcomes) (n = 2)
Results Participant flow and recruitment Sixty-nine patients (40 interventions, 29 controls) were recruited between April 2012 and March 2013. One patient has been excluded after randomization, as he was found not to meet inclusion criteria (HbA1c \ 7.5 %). Sixty-eight patients completed 3 months and 65 patients completed the entire 6 months study (Fig. 1).
Baseline data Both groups had similar baseline demographic and clinical characteristics, including glycemic control and lipid levels, except for significantly higher mean HDL levels in the intervention group [values in both groups
Int J Clin Pharm Table 1 Baseline characteristics of study patients Characteristics
Interventiona
Controla
Total number of patients (n)
40
28
Bosnia-Herzegovina (n)
26
20
0.7685*
Germany (n)
14
8
0.7685*
Gender [n (%)] Female
24 (60)
15 (54)
0.7807*
Age (years), (range)
14.5 (12–17)
14.9 (12–18)
0.2874**
Body Mass Index (BMI), (range)
20.8 (14.5–27.5)
21.7 (17.1–29.1)
0.1864**
Duration of diabetes (years), (range)
5.9 (1–14)
6.8 (1–14)
0.2317**
Patients with 1–3 comorbid conditions [n (%)]
16 (40)
15 (54)
0.3751*b
Average number co-medications [n (range)]
1.1 (0–2)
1.3 (0–3)
0.6088*
9.4
9.4
0.5833*
[7.5 to \9 % [n (%)]
19 (48)
14 (50)
0.8391*
C9 % [n (%)]
21 (53)
14 (50)
0.8391*
P value
Glycemic parameters HbA1c [%] HbA1c categories
Blood glucose (BG) mg/dL, (range)
224.3 (30.6–555.0)
215.4 (28.8–470.3)
0.9404*
Home BG tests per day (n), (self-reported)
4.2
4.0
0.8885#
Insulin (IU/day), (range)
53.7 (32–80)
55.4(29–81)
0.6809**
Insulin pen (multiple daily injections) [n (%)]
24 (60)
14 (50)
0.5692*
Insulin pump (CSII) [n (%)]
16 (40)
14 (50)
0.5692*
Systolic (mmHg), (range)
118 (100–134)
121(105–140)
0.2054**
Diastolic (mmHg), (range)
71 (51–98)
74 (46–100)
0.3263**
171.4 (104–282)
178.4 (108–344)
0.6092**
Low density lipoprotein (LDL) (mg/dL), (range)
104.2 (62–164)
108.4 (63–230)
0.8811**
High density lipoprotein (HDL) (mg/dL), (range) Triglycerides (mg/dL), (range)
65.2 (16–124) 115.7 (25–1128)
54.7 (18–139) 132.8 (9–703)
0.001* 0.6743*
Sum of incidences of diabetic Ketoacidosis(DKA)
12
14
0.1411#
Sum of incidences of severe hypoglycaemia
8
6
1.000#
Drug therapy
Blood pressure (BP)
Biochemical parameters Total cholesterol (mg/dL), (range)
Acute complications (time period: 6 months before study)
a
Mean values are reported, unless otherwise indicated
b
It was tested whether the number of co-morbid conditions depends on the group membership
* Chi squared test ** Mann–Whitney test #
Fisher’s exact test
are considered clinically normal ([35 mg/dL)] [32] (Table 1). Co-morbidities were reported in 36.8 % of patients (41 % in Bosnia and 26 % in Germany). The most common co-morbidities were hypothyroidism in Bosnia (patients were stabilized on L-Thyroxine 25–50 mcg before the study), and asthma, hay fever, and lactose intolerance in Germany. There were no significant between-group differences in the incidences of diabetic ketoacidosis and severe hypoglycaemia in the previous 6 months (Table 1).
Pharmaceutical care intervention In total, 38 patients in the intervention group completed C6 face-to-face PhC, 1 patient completed 5 face-to-face PhC visits. In majority of patients study assessments were performed by documenting C6 CRFs per patient and evaluating C6 BG records. At least 6 PCPs were written per patient, containing C6 goals, and C2 PCPs (visits 3 and 6) were submitted to physicians. Five follow-ups per patient were performed by the pharmacists.
123
Change from baseline in HbA1c (%)
0.4281 0.038
0.0001 0.0122
0.4255
0.0030
0.0075 0.00002 0.0153
6 months versus baseline 3 months versus baseline 6 months
6 months
3 months *p = 0.00002
0.6
*p = 0.0075
0.4 0.2 0 -0.2 -0.4 -0.6 -0.8
0.2450
Control n = 28
Intervention n = 39
Control n = 26
Primary outcome Relative to usual care, the PhC intervention reduced HbA1c. Mean HbA1c decreased from 9.4 % at baseline to 8.9 % at 6 months in the intervention group, compared with an increase from 9.4 to 9.9 % in the control group (change from baseline -0.54 vs. ?0.32 %, p = 0.0075) (Table 2; Fig. 2). The between-group-difference was greatest after 3 months (-1.09 vs. ?0.23 %, p = 0.00002); (Table 2; Fig. 2). Mean HbA1c reductions were numerically (but not statistically) greater in poorly controlled patients (baseline HbA1c [9 %) than in moderately controlled patients (baseline HbA1c C7.5 to \9 %) after 3 (p = 0.050) and 6 months (p = 0.6978). When analysed by country, a significantly greater mean HbA1c improvement in the intervention group versus the control group was observed at both 3 and 6 months in Bosnia-Herzegovina, but only at 3 months in Germany (all p \ 0.04) (Table 2). Because HbA1c outcomes differed between countries, (potentially influencing the overall HbA1c outcome), linear regression was performed to adjust for the potential country effect, showing that the factor ‘‘country’’ had no effect on the differences in HbA1c between baseline and 6 months (p = 0.5879); only the factor ‘‘group’’ had a significant influence (p = 0.0078). Severe hypoglycemic events ** Mann–Whitney test
HbA1c values between intervention and control at different time points: baseline, after 3 months and after 6 months
Difference in HbA1c values: intervention versus control after 3 months relative to baseline and after 6 months relative to baseline
Intervention n = 40
Fig. 2 HbA1c % difference in the intervention and control groups after 3 months versus baseline and after 6 months versus baseline
c
Data are mean values, unless otherwise indicated a
n, number of patients; HbA1c, Glycosylated haemoglobin
9.1 (7.1–12.5)
9.9 (7.7–12.9) 8.9 (7.2–11.4) 9.5 (7.5–12.2) Germany (range)
9.3 (7.5–14.7) BosniaHerzegovina (range)
By country
8.8 (6.6–13.1)
8.9 (6.6–13.1) Total HbA1c (%) (range)
b
0.5160 9.7 (6.8–11.7)
0.4178 10.0 (7.7–13.0) 9.6 (7.2–12.8) 8.0 (5.7–11.2) 9.3 (7.5–13.6)
9.4 (7.5–14.7) 8.3 (5.7–11.4) 9.4 (7.5–13.6)
123
*Mann Whitney Test
9.8 (7.8–12.8)
0.0037
0.0020 0.5833 9.9 (6.8–13.0) 9.7 (7.2.–12.8)
Baseline Baseline (n = 28) 3 months (n = 40) Baseline (n = 40)
6 months (n = 39)
Control groupa Intervention groupa
3 months (n = 28)
6 months (n = 26)
p value**c
3 months
-1.0
Parameters
Table 2 HbA1c flow and HbA1c difference after 3 and 6 months in the Intervention and Control Groups in the Total Population and by Country
p value** of HbA1c differences between groupsa,b
Int J Clin Pharm
There were no significant between-group differences in the number of severe hypoglycemic events during the study (p = 0.1276) or in the previous 6 months (p = 1). During the study, one severe hypoglycemic event was reported by 7 of 40 patients in the intervention group and 1 of 28 patients in the control group. In the previous 6 months, one
Int J Clin Pharm
event was reported in 6 and 4 patients in the respective groups, while two events were reported in one patient in each group. Well-being Index (WHO-5) (intervention group) In the intervention group, baseline mean well-being score was 52.8 % (range 16–84 %); however, 45 % of patients scored \50 % (indicating low mood) and 13 % scored \28 % (indicating likely depression) [30]. Wellbeing scores improved significantly to 59.2 % (range 24–100 %) at 3 months (p = 0.0002) and to 63.3 % (range 24–96 %) at 6 months (p = 0.00002). At 3 months, 33 and 3 % of patients scored \50 and \28 %, respectively; the corresponding percentages at 6 months were 26 and 3 %. Moreover, the change from baseline in well-being at 6 months was [10 %, which is regarded as clinically relevant [30]. Linear regression revealed that the covariates country (p = 0.9563), age (p = 0.7033), insulin device (p = 0.7169), duration of T1DM (p = 0.9850), and hypoglycaemia at baseline (p = 0.2337) had no influence on the wellbeing scores after 6 months.
Discussion This study demonstrates that PhC in adolescents with T1DM and poor glycemic control contributed glycemic control, as shown by significant HbA1c reductions with no increase in severe hypoglycaemia. Our findings however, in particular the HbA1c reduction, are in contrast to previous research [15, 16]. The lack of benefit in these studies may result from the absence of a structured PhC [17, 18]. In our study, pharmacists assessed clinical data and developed PCPs with problem-solving interventions in concert with the patients to define responsibilities [18] and improve adherence. The PCPs were discussed and progress evaluated with physicians. Therefore, implementation of an appropriate, structured, and multidisciplinary PhC [17, 18], along with patient empowerment, seems crucial for the improved glycemic control we observed. Large studies (n = 289 and 288) in Australia and Belgium with [ 50 community pharmacies evaluating PhC impact on clinical outcomes in adults with T2DM also reported HbA1c reductions (ranging 0.5–1.0 %) [33, 34]. Thus, our trial further supports the value of pharmacists in a multidisciplinary team to reduce disease burden in T1DM. The well-being perception in our study improved by [10 % after 6 months, which is considered to be clinically relevant [30]. However, because well-being was evaluated only in the intervention group, we sought to compare our results with those from other studies as a reference; two studies that included pharmacist
intervention [35, 36] showed similar results to ours. A study by Petkova et al. [35] reported an improvement of 11.7 % in patients with T1DM, and a study by Fresenius [36] reported an improvement of [10 % in patients with T1DM and T2DM. These findings, which are comparable with our results, indicate that PhC is likely to have a positive influence on patient well-being, as the wellbeing in patients who received clinical care from pharmacists did not worsen over the course of the studies. The simple coin-toss randomization method was selected in our study for practical, organizational and coordination reasons, as the same method needed to be used in both study centers. Given the small sample size, this simple randomization resulted in unequal numbers of patients in the study groups. However, unequal randomization ratios significantly reduce study power only if the ratio is 3:1 or more [37–39], which is not expected in our study with an apparent ratio of 4:3. Most importantly, our simple randomization resulted in well balanced study groups (Table 1). Remarkably, in both countries, a greater HbA1c reduction was achieved after 3 rather than 6 months, possibly because of decreasing patient motivation. This finding supports previous studies [6, 16, 40] showing adherence challenges in adolescents (intrinsic to the developmental stage [6]) and requires a longer follow-up with new approaches to sustain the initial improvement. Although HbA1c reduction was achieved in both countries, the effect was more prominent in Bosnia-Herzegovina, which is probably due to the clinical pharmacist’s on-site presence, leading to more frequent and a more intensive patient-pharmacist-physician contact. As a matter of fact, more patients in Bosnia received additional PhC visits and calls then in Germany and as is known, more frequent visits are associated with glycemic control [41–43]. Also, in Bosnia, patients start from a lower baseline of care, and any patient-focused approach is likely to yield a substantial improvement. Limited health care and overall resource shortage may have contributed to better acceptance and adherence by the youth in Bosnia. Acute and long-term T1DM complications result in considerable premature mortality [9, 44]. The Epidemiology of Diabetes Interventions and Complications (EDIC) study confirmed that lowering HbA1c by 1.7 % (HbA1c difference 8.1 vs. 9.8 %) decreases the risk of retinopathy by 53 %, microalbuminuria by 54 %, and neuropathy by 60 % [44]. Therefore, the HbA1c reduction of 0.54 % in our study has potential clinical relevance, particularly in high-risk, poorly controlled patients, who may benefit from the PhC, more frequent visits [41–43] and extended support by the pharmacists. Limitations There are limitations to our study. A major limitation is the heterogeneous study settings between the two countries—
123
Int J Clin Pharm
single clinical pharmacist in Bosnia combined with community pharmacists in Germany which limits study’s external and internal validity. Furthermore, a priori sample size calculation was not performed. All available patients in the clinical setting were screened and incorporated. A post hoc power analysis, however, was performed indicating a study power of 81.90 %. The study results were analyzed per protocol and not by intention to treat. A simple coin-toss randomization method resulted in unequal study groups, as the participant numbers were too small to achieve a relatively even split between comparator groups. A larger number of patients investigated in the same practice environment may be associated with a higher validity than the setting in the DIADEMA trial. The short duration implies that the results may become less favorable with a longer study duration, due to decreasing patient motivation [6, 40], warranting further research on methods to sustain the PhC improvements. Due to time constraints, secondary parameters such as well-being was recorded only for the intervention group. Finally, two study sites in only two countries are unlikely to represent the general adolescent T1DM population; healthcare differences (including treatment and counselling) may pose barriers to PhC implementation in other settings [45].
Conclusion Our study suggests that multidisciplinary PhC may add value in the management of T1DM in adolescents and inadequate glycemic control, as shown by HbA1c reduction. Integration of pharmacists into multidisciplinary T1DM teams may potentially help reduce the rate of complications of diabetes and improve existing care structures across multiple healthcare systems. However, the optimal methods including homogenous settings, a lager cohort and a longer follow-up, on how to achieve sustained, long-term improvements require further study. Acknowledgments The authors thank the patients. They also thank the participating pharmacists (in alphabetical order): Henrich Dieter Backes, Zora Bahser, Hans-Dieter Brink, Dieter Conze, Kathrin Furth, Klaus Dieter von Laguna, Anja Mu¨ller, Norbert Mu¨ller, Anne Rhein, Mara Scholz, Friederike Sieben, Ralf Weckop, Katja Weissenborn, and the pediatric clinic directors: Prof. Dr. med. Tim Niehues and Prof. Dr. med. Senka Mesihovic-Dinarevic. We are indebted to K. A. Lyseng-Williamson and L.Yang for editorial assistance. Funding Emina Obarcanin was financially supported by the Deutsche Akademische Auslandsdienst (DAAD), the Lesmu¨ller Stiftung, and the Heinrich-Heine University, Du¨sseldorf, Germany. The study was partially supported by the Lesmu¨ller Stiftung. Conflicts of interest
123
The authors have no conflicts of interest.
References 1. The DIAMOD project group. Incidence and trends of childhood Type 1 diabetes worldwide 1990–1999. Diabet Med. 2006;23:857–66. 2. Patterson C, Dahlquist G, Gyu¨ru¨s E, Green A, Solte´sz G. EURODIAB Study Group. Incidence trends for childhood type 1 diabetes in Europe during 1989–2003 and predicted new cases 2005–20: a multicentre prospective registration study. Lancet. 2009;373:2027–33. 3. Ehehalt S, Dietz K, Willasch AM, Neu A, Baden-Wu¨rttemberg Diabetes Incidence Registry (DIARY) Group. Epidemiological perspectives on type 1 diabetes in childhood and adolescence in Germany. Diabetes Care. 2010;33:338–40. 4. Tahirovic H, Toromanovic A. Incidence of type 1 diabetes mellitus in children in Tuzla Canton between 1995 and 2004. Eur J Pediatr. 2007;166:491–2. 5. Daneman D, Hamilton J. Is poor metabolic control inevitable in adolescents with type 1 diabetes? An Pediatr. 2001;1:40–4. 6. Borus J, Laffel L. Adherence challenges in the management of type 1 diabetes in adolescents: prevention and intervention. Curr Opin Pediatr. 2010;22:405–11. 7. Patino AM, Sanchez J, Eidson M, Delamater AM. Health beliefs and regimen adherence in minority adolescents with type 1 diabetes. J Pediatr Psychol. 2005;30:503–12. 8. Schober E, Wagner G, Berger G, Gerber D, Mengl M, Sonnenstatter S, et al. Prevalence of intentional under-and overdosing of insulin in children and adolescents with type 1 diabetes. Pediatr Diabetes. 2011;12:627. 9. Harjutsalo V, Forsblom C, Groop P. Time trends in mortality in patients with type 1 diabetes: nationwide population based cohort study. BMJ. 2011;343:5364. 10. Petersen M. Economic costs of diabetes in the U.S. in 2012. Diabetes Care. 2013;36:1033–46. 11. Daneman D. Type 1 diabetes. Lancet. 2006;367:847–58. 12. Wermeille J, Bennie M, Brown I, McKnight J. Pharmaceutical care model for patients with type 2 diabetes: integration of the community pharmacist into the diabetes team-a pilot study. Pharm World Sci. 2004;26:18–25. 13. Chisholm-Burns MA, Kim Lee J, Spivey CA, Slack M, Herrier RN, Hall-Lipsy E, et al. US pharmacists’ effect as team members on patient care: systematic review and meta-analyses. Med Care. 2010;48:923–33. 14. Machado M, Bajcar J, Guzzo GC, Einarson TR. Sensitivity of patient outcomes to pharmacist interventions. Part I: systematic review and meta-analysis in diabetes management. Ann Pharmacother. 2007;41:1569–82. 15. Gay CL, Chapuis F, Bendelac N, Tixier F, Treppoz S, Nicolino M. Reinforced follow-up for children and adolescents with type 1 diabetes and inadequate glycaemic control: a randomized controlled trial intervention via the local pharmacist and telecare. Diabetes Metab. 2006;32:159–65. 16. Sims LM, Haines SL. Challenges of a pharmacist-directed peer support program among adolescents with diabetes. J Am Pharm Assoc. 2011;51:766–76. 17. Cipolle RJ, Strand LM, Morley PC. Pharmaceutical care practice: The clinician’s guide. 2nd ed. New York: McGraw-Hill; 2004. 18. van Mil JW, Schulz M, Tromp TF. Pharmaceutical care, European developments in concepts, implementation, teaching, and research: a review. Pharm World Sci. 2004;26(6):303–11. 19. Bosnia-Herzegovina Agency for Drugs and Medical Devices. Clinical trials [Agencija za lijekove i medicinska sredstva Bosne i Hercegovine]. http://www.almbih.gov.ba/klinicka-ispitivanja. Accessed 19 Sep 2014.
Int J Clin Pharm 20. Bosnia-Herzegovina Agency for Drugsand Medical Devices. Clinical trial and medical devices directive of Bosnia and Herzegovina. Official Gazette of Bosnia and Herzegovina No. 4/10 [Sluzbeni Glasnik Bosne i Hercegovine Nr. 4/10 http:// www.almbih.gov.ba/_doc/regulative/pravilnik_klinicka_bos.pdf. Accessed 19 Sep 2014. 21. Ineternational Society of Pediatric Diabetes. Clinical practice consensus guidelines 2009. http://www.ispad.org/content/ispadclinical-practice-consensus-guidelines-2009. Accessed 19 Sep 2014. 22. Faul F, Erdfelder E, Lang A-G, Buchner A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39:175–91. 23. Blume J, Peipert JF. Randomization in controlled clinical trials: why the flip of a coin is so important. J Am Assoc Gynecol Laparosc. 2004;11:320–5. 24. Berger W. A review of methods for ensuring the comparability of comparison groups in randomized clinical trials. Rev Recent Clin Trials. 2006;1:81–6. 25. Clark M, Westerberg B. How random is the toss of a coin? CMAJ. 2009;181:306–8. 26. Viera A, Bangdiwala I. Eliminating bias in randomized controlled trials: Importance of allocation concealment and masking. Fam Med. 2007;39:132–7. 27. German National Treatment Guideline Program. Diabetes structured education program Version 3 Jun 2013 [Program fuer nationale Versorgungslinie, Diabetes strukturierteSchulungsprogramme].http:// www.versorgungsleitlinien.de/themen/diabetes2/dm2_schulung/ pdf/nvl-t2d-schulung-lang-3.pdf. Accessed 19 Sep 2014. 28. Workgroup on Hypoglycemia, American Diabetes Association. Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia. Diabetes Care. 2005;28(5):1245–9. 29. De Wit M, Pouwer F, Gemke RJ, Delemarre-van de Waal HA, Snoek FJ. Validation of the WHO-5 Well-Being Index in adolescents with type 1 diabetes. Diabetes Care. 2007;30:2003–6. 30. WHO (Five) Well-Being Index (1998 version). Hillerod: Psychiatric Research Unit, WHO Collaborating Center for Mental Health, Frederiksborg General Hospital. 1998. http://www.psy kiatri-regionh.dk/NR/rdonlyres/ACF049D6-C94D-49B2-B34C96A7C5DA463B/0/WHO5_English.pdf. Accessed 19 Sep 2014. 31. R Foundation for Statistical Computing 2014. A Language and Environment for Statistical Computing. R Core Team Vienna, Austria: http://www.r-project.org. Accessed 19 Sep 2014.
32. American Diabetes Association. Management of dyslipidemia in children and adolescents with diabetes. Diabetes Care. 2003;26:2194–7. 33. Krass I, Armour CL, Mitchell B, Hughes J, Peterson G, Song YJ, et al. The Pharmacy Diabetes Care Program: assessment of a community pharmacy diabetes service model in Australia. Diabetic Med. 2007;24:677–83. 34. Mehuys E, Van Bortel L, De Bolle L, Van Tongelen I, Annemans L, Remon JP. Effectiveness of a community pharmacist intervention in diabetes care: A randomized controlled trial. J Clin Pharm Ther. 2011;36:602–13. 35. Petkova V, Ivanova A, Petrova G. Education of patients with diabetes in the community pharmacies (pilot project in Bulgaria). Ankara Ecz Fak Derg. 2006;35:111–24. 36. Fresenius K. Impact of pharmaceutical care on clinical and therapy outcomes in patients with diabetic foot syndrom (DFS) [Dissertation]. Institute for Chemistry, Pharmacy and Geo-Science, University of Mainz, Germany: Fresenius K, 2007. 37. Pocock SJ. Clinical trials: a practical approach. Chichester: Wiley; 1995. 38. Gail M, Williams R, Byar DP, Brown C. How many controls? J Chronic Dis. 1976;29:723–31. 39. Dumville JC, Hahn S, Miles JN, Torgerson DJ. The use of unequal randomisation ratios in clinical trials: a review. Contemp Clin Trials. 2006;27:1–12. 40. Frank M. Psychological issues in the care of children and adolescents with type 1 diabetes. Paediatr Child Health. 2005;10:18–20. 41. Phan TL, Hossain J, Lawless S, Werk NL. Quarterly visits with glycated hemoglobin monitoring: The sweet spot for glycemic control in youth with type 1 diabetes. Diabetes Care. 2014;37:341–5. 42. Kaufman FR, Halvorson M, Carpenter S. Association between diabetes control and visits to a multidisciplinary pediatric diabetes clinic. Pediatrics. 1999;103:948–51. 43. Svoren BM, Butler D, Levine BS, Anderson BJ, Laffel LM. Reducing acute adverse outcomes in youths with type 1 diabetes: A randomized, controlled trial. Pediatrics. 2003;112:914–22. 44. The Epidemiology of Diabetes Interventions and Complications (EDIC) study. Sustained effect of intensive treatment of type 1 diabetes mellitus on development and progression of diabetic nephropathy. JAMA 2003; 2159–2167. 45. van Mil JW, de Boer W, Tromp TH. European barriers to the implementation of pharmaceutical care. International Journal of Pharmacy Practice. 2001;9:163–8.
123
RESEARCH ARTICLE
Pharmaceutical care of adolescents with diabetes mellitus type 1: the DIADEMA study, a randomized controlled trial Emina Obarcanin1 • Manfred Kru¨ger2 • Petra Mu¨ller3 • Verena Nemitz3 • Holger Schwender4 • Snijezana Hasanbegovic5 • Sena Kalajdzisalihovic5 • Stephanie La¨er1
Received: 22 June 2014 / Accepted: 14 April 2015 Koninklijke Nederlandse Maatschappij ter bevordering der Pharmacie 2015
Abstract Background Physiological and psychological changes during puberty and a low adherence to complex treatment regimens often result in poor glycemic control in adolescents with type 1 diabetes mellitus (T1DM). The benefit of pharmaceutical care in adults with diabetes mellitus type 2 has been explored; however, evidence in adolescents with T1DM is scarce. Objective To evaluate the impact of pharmaceutical care in adolescents with T1DM provided by pharmacists, in collaboration with physicians and diabetes educators on important clinical outcomes (e.g., HbA1c and severe hypoglycemia) Setting: At the outpatient Helios Paediatric Clinic and at the 12 regular community pharmacies of the study patients with 14 pharmacists in the Krefeld area, Germany, and at the University Pediatric Clinic with one clinical pharmacist on-site in Sarajevo, Bosnia-Herzegovina. Methods A randomized, controlled, prospective, multicenter study in 68 adolescents with T1DM. The intervention group received monthly structured pharmaceutical care visits delivered by pharmacists plus supplementary visits and phone calls on an as needed basis, for 6 months. The control group received usual diabetic care. Data were collected at baseline and after 3 and 6 months. Main outcome measures: The
between-group difference in the change from baseline in glycosylated hemoglobin (HbA1c) and the number of severe hypoglycemic events in both groups. Results The improvement from baseline in HbA1c was significantly greater in the intervention group than in the control group after 6 months (change from baseline -0.54 vs. ?0.32 %, p = 0.0075), even after adjustment for country-specific variables (p = 0.0078). However, the effect was more pronounced after only 3 months (-1.09 vs. ?0.23 %, p = 0.00002). There was no significant between-group difference in the number of severe hypoglycemia events. (p = 0.1276). Conclusion This study suggests that multidisciplinary PhC may add value in the management of T1DM in adolescents with inadequate glycemic control. However, the optimal methods on how to achieve sustained, long-term improvements in this challenging population require further study. Keywords Adolescents Bosnia-Herzegovina Germany Glycemic control Pharmaceutical care Pharmacists Type 1 diabetes mellitus
Impacts on Practice & Emina Obarcanin [email protected] 1
Department of Clinical Pharmacy and Therapeutics, Heinrich Heine University, Universita¨tsstrasse 1, 40225 Du¨sseldorf, Germany
2
Linner Pharmacy, Krefeld, Germany
3
Helios Pediatric Clinic, Krefeld, Germany
4
Mathematical Institute, Heinrich Heine University, Du¨sseldorf, Germany
5
Pediatric Clinic, University Clinical Center Sarajevo, Sarajevo, Bosnia-Herzegovina
•
•
•
Pharmaceutical care may contribute to the improvement of glycemic control in adolescents with type 1 diabetes mellitus. Implementation of pharmaceutical care in a multidisciplinary team may help to prevent acute and, potentially, long-term complications of T1DM. The pharmaceutical care protocol developed for adolescents with T1DM may help to bridge the gap in healthcare provision across different healthcare systems, especially in countries where healthcare resources are limited.
123
Int J Clin Pharm
Introduction Type 1 diabetes mellitus (T1DM) incidence in children and adolescents is increasing worldwide [1–4]. Physiological changes with greater insulin resistance during puberty, as well as psychological factors and non-adherence to insulin regimens, result in poor glycemic control [5–8], leading to increased morbidity, and premature mortality [5, 9]. T1DM complications represent a major economic burden [10]. A multidisciplinary approach with intensive patient monitoring is required to prevent complications of diabetes and improve quality of life [11]. Multidisciplinary pharmaceutical care (PhC) programs have improved outcomes in adults with type 2 diabetes mellitus (T2DM) [12–14]. However, data on such programs in adolescents with T1DM remain scarce. A randomized, controlled, French study with pharmacists faxing glucometer data to the hospital revealed no significant difference in HbA1c between the study groups [15]; an observational American study in which pharmacists directed a peer-support program showed even an increase in HbA1c [16]. These findings indicate that pharmacist interventions alone, in the absence of PhC (as defined in Standards of Practice [17]) are not sufficient to improve outcomes in this population. Therefore, further study was needed to evaluate the effects of integrating a structured PhC program [17] in adolescents with T1DM. Our DIADEMA (DIabetes in ADolescence Engagement and Monitoring by phArmacists) study, designed according to Standards of Practice [17], incorporated clear responsibility assignment, counseling, and outcome evaluation [18]. This involved: (1) assessment of drug-related needs through data collection and interpretation, (2) development of Pharmaceutical Care Plans (PCPs) with problem-solving interventions and individual goals, (3) follow-up and outcome evaluation. We hypothesized that implementing this multidisciplinary, structured PhC would improve HbA1c and other clinical outcomes.
Aim of the study We evaluated the impact of intense and structured diabetesrelated pharmaceutical care on clinical outcomes in adolescents with T1DM in Germany and Bosnia-Herzegovina. Ethical approval The study was approved by the Ethics Committee of the Medical Faculty of Heinrich-Heine University, Du¨sseldorf, Germany (No. 3991). The Agency for Medicinal Products and Medical Devices of Bosnia-Herzegovina and the Ethical Committee of the Clinical Center at the University
123
of Sarajevo did not require ethical approval, as they did not qualify this study as a clinical trial of drugs (Reference No. 10-07-56-1207/12), (Official Gazette Nr. 4/10) [19, 20]. The study was performed according to the Declaration of Helsinki and Good Clinical Practice (GCP) guidelines. Study patients were verbally informed about the study and patients and their parents signed written consent prior to the start of the study.
Methods Study design This was a randomized, controlled, prospective 6-month trial conducted in adolescents with T1DM and poor glycemic control. Participants and setting Eligible patients (12–18 years of age) who had confirmed T1DM diagnosed C6 months prior to enrolment were outside the partial remission phase, as indicated by insulin use of C0.5 IU/kg/day [21], used intensified conventional therapy (ICT) or continuous subcutaneous insulin infusion (CSII) and had C2 measurements of HbA1c consistently C7.5 %, taken C3 months apart. Exclusion criteria included intellectual and developmental disabilities, drug abuse, psychiatric conditions, pregnancy, and breastfeeding. The study had two sites: Pediatric Clinic Helios Hospital Krefeld in Germany (with 12 community pharmacies and 14 pharmacists) and University Pediatric Clinic Sarajevo in Bosnia-Herzegovina (with one on-site clinical pharmacist). Recruitment and randomization The clinics recruited eligible patients. Patients in Germany named their regular community pharmacies, from which 14 pharmacists were recruited. A post hoc power analysis based on the actual sample and effect size was performed using GPower version 3.1 [22]. Randomization was performed by diabetes educators (nurses) using a simple, easy and well analyzed coin-toss method [23–25]. Physicians supervised the coin toss to avoid potential bias [24, 25]. Nurses wrote eligible patient names and sealed these in envelopes to conceal allocation and minimize selection bias [24, 26]. The coin toss resulted in 40 ‘‘tails’’ (predesignated as the intervention group) and 29 ‘‘heads’’ (control). Following randomization, envelopes in each group were consecutively numbered by nurses and opened. To improve willingness to participate, patients who
Int J Clin Pharm
consented verbally were randomized and then signed consent prior to the study start.
study coordinators (M. Kru¨ger and E. Obarcanin) on PhC, T1DM management and GCP prior to and during the study.
Data collection
Control group
Diabetes educators (nurses) and pharmacists collected and exchanged data using two-part CRFs. Participants attended the clinics 3 times (baseline and 3 and 6 months) to receive usual care (usually every 3 months), including HbA1c measurement and counseling by the diabetes team. Nurses recorded socio-demographic and other data (e.g., HbA1C, insulin dose) at each visit, using ‘‘Physician CRFs’’; these were verified by physicians, according to GCP. The CRFs of the intervention group were accessible on-site in Bosnia and were faxed to the pharmacists in Germany.
Participants had scheduled diabetic clinic appointments every 3 months. As part of the ‘‘usual care visits’’ in the clinics, patients were advised to measure their BG at least 4 times daily, [21] and the frequency of BG monitoring was routinely documented by diabetes educators (nurses). During the study, patients could request counseling from their local pharmacists in special cases (e.g. school excursions, hypoglycemia). However, as these local pharmacists were not study participants, we could not confirm if counseling was requested or received.
Intervention group
Primary outcome
Patients received monthly (over 6 months) scheduled PhC visits (60-90 min) with pharmacists, plus on-demand phone calls and additional visits on an as needed basis, as required by patients and in special situations (e.g., repeated uncontrolled hyperglycemia, school excursions, frequent hypoglycemia). The PhC was standardized in both countries according to the Standards of Practice [18]. During the visits pharmacists recorded clinical data (i.e., blood glucose (BG), hypoglycemia, ketone values, insulin therapy and dose, tolerability) using ‘‘Pharmacist CRFs’’. Patients in the intervention group measured and documented their BG at least 4 times daily (i.e., fasting, after administering insulin, bedtime) [21], preferably 5–6 times to include measurements 2 h after meals or during or after sports. Based on data and evaluation of BG records [21], pharmacists assessed drug-related needs and identified problems (e.g., insulin non-adherence at school, dose adjustment for hyperglycemia or sports). Individual Pharmaceutical Care Plans (PCPs) were developed for each patient, including problem-solving interventions, (e.g. dose adjustments, diet modification, contacting teachers) and C1 measurable goal (e.g., HbA1c reduction in 3 months). Immediate insulin dose changes pharmacists proposed to and discussed with physicians personally (in Bosnia) or by phone (in Germany). Every PCP of visits 3 and 6 had to be faxed/submitted and discussed with physicians. In Germany, multidisciplinary face-to face meetings between the community pharmacists, physician and diabetes educator took place before and at the study start and every 2 months during the study. Pharmacists also provided education on standardized T1DM-related topics: treatment goals, insulin use, self-monitoring of blood glucose (SMBG), prevention of complications, and physical activity [27]. Follow-up and progress evaluation was performed monthly. Individual and group training were provided to pharmacists by the
The primary endpoint was the change in HbA1c from baseline at 3 and 6 months. Secondary outcomes (both groups) Mean number of severe hypoglycemic events, defined as low glucose requiring assistance of another person [28], was recorded by nurses from multiple sources (e.g., medical records, hospitalizations, BG records, and glucometer data) and was retrospectively analyzed for the 6 months prior to the study, and during the study. Well-being Index (WHO-5) (intervention group) A validated instrument was used to assess psychological status [29, 30] at months 3 and 6. The five positively worded items (i.e., ‘‘I have felt: cheerful spirits; calm; active; I woke up fresh; my day is interesting’’) measure well-being over the past 2 weeks using a 6-point Likert scale from 0 (not present) to 5 (constantly present).Total scores were multiplied by 4 to obtain percentage scores of 0–100, with higher scores indicating better well-being. Scores \50 and \28 % indicate low mood and depression, respectively. Repeated scores of \28 % were conveyed to the patient’s physician, with a clinical psychologist consult if necessary. Data analysis Analysis of data was performed in the per-protocol population (Fig. 1). Statistical analysis, including baseline characteristics and HbA1c data collected at baseline and after 3 and 6 months, was performed using the statistical software environment R, version 3.0.1 [31]. The Mann– Whitney test for quantitative (non-normally distributed)
123
Int J Clin Pharm
Screening
Assessed for eligibility (n = 80)
Not included (n =11) Not meeting inclusion criteria (n =3) Declined to participate (n =8)
Allocation
Randomization
Fig. 1 Patient flowchart of the DIADEMA trial
Randomized (n = 69)
Allocated to pharmaceutical care (n = 40) Bosnia (n = 26); Germany (n = 14)
Analysis
Follow-up
Received allocated intervention (n = 40)
Discontinued intervention (n =1) Declined to participate after 3 months due to lack of motivation
Pharmaceutical care group Analyzed (n =39) Excluded from analysis (study withdrawal) (n =1)
variables was used to analyze primary and secondary outcomes, and the v2 or Fisher’s exact tests were used to analyze categorical variables in the secondary outcomes. The p-values were calculated and used to interpret the statistical significance of the differences between the groups. For the primary outcome analysis, the Mann–Whitney test was used, as the differences in HbA1c values at different time points were non-normally distributed or only marginally normally distributed. Furthermore, a linear regression analysis and analysis of variance with the Wald test were applied to test whether the variables ‘‘country (Bosnia/Germany) and ‘‘Group’’ (intervention/control) significantly influenced HbA1c changes. Further analysis of the well-being (WHO-5), included linear models to test if baseline covariates such as age, diabetes duration, sex, and insulin device influenced the results.
123
Allocated to usual care (n =29) Bosnia (n=20); Germany (n =9) Excluded, not meeting inclusion criteria (n =1) Finally allocated to usual care (n =28)
Lost to follow up (n =2) Did not come to the control visit after 6 months
Usual care group Analyzed (n =26) Excluded from analysis (missing outcomes) (n = 2)
Results Participant flow and recruitment Sixty-nine patients (40 interventions, 29 controls) were recruited between April 2012 and March 2013. One patient has been excluded after randomization, as he was found not to meet inclusion criteria (HbA1c \ 7.5 %). Sixty-eight patients completed 3 months and 65 patients completed the entire 6 months study (Fig. 1).
Baseline data Both groups had similar baseline demographic and clinical characteristics, including glycemic control and lipid levels, except for significantly higher mean HDL levels in the intervention group [values in both groups
Int J Clin Pharm Table 1 Baseline characteristics of study patients Characteristics
Interventiona
Controla
Total number of patients (n)
40
28
Bosnia-Herzegovina (n)
26
20
0.7685*
Germany (n)
14
8
0.7685*
Gender [n (%)] Female
24 (60)
15 (54)
0.7807*
Age (years), (range)
14.5 (12–17)
14.9 (12–18)
0.2874**
Body Mass Index (BMI), (range)
20.8 (14.5–27.5)
21.7 (17.1–29.1)
0.1864**
Duration of diabetes (years), (range)
5.9 (1–14)
6.8 (1–14)
0.2317**
Patients with 1–3 comorbid conditions [n (%)]
16 (40)
15 (54)
0.3751*b
Average number co-medications [n (range)]
1.1 (0–2)
1.3 (0–3)
0.6088*
9.4
9.4
0.5833*
[7.5 to \9 % [n (%)]
19 (48)
14 (50)
0.8391*
C9 % [n (%)]
21 (53)
14 (50)
0.8391*
P value
Glycemic parameters HbA1c [%] HbA1c categories
Blood glucose (BG) mg/dL, (range)
224.3 (30.6–555.0)
215.4 (28.8–470.3)
0.9404*
Home BG tests per day (n), (self-reported)
4.2
4.0
0.8885#
Insulin (IU/day), (range)
53.7 (32–80)
55.4(29–81)
0.6809**
Insulin pen (multiple daily injections) [n (%)]
24 (60)
14 (50)
0.5692*
Insulin pump (CSII) [n (%)]
16 (40)
14 (50)
0.5692*
Systolic (mmHg), (range)
118 (100–134)
121(105–140)
0.2054**
Diastolic (mmHg), (range)
71 (51–98)
74 (46–100)
0.3263**
171.4 (104–282)
178.4 (108–344)
0.6092**
Low density lipoprotein (LDL) (mg/dL), (range)
104.2 (62–164)
108.4 (63–230)
0.8811**
High density lipoprotein (HDL) (mg/dL), (range) Triglycerides (mg/dL), (range)
65.2 (16–124) 115.7 (25–1128)
54.7 (18–139) 132.8 (9–703)
0.001* 0.6743*
Sum of incidences of diabetic Ketoacidosis(DKA)
12
14
0.1411#
Sum of incidences of severe hypoglycaemia
8
6
1.000#
Drug therapy
Blood pressure (BP)
Biochemical parameters Total cholesterol (mg/dL), (range)
Acute complications (time period: 6 months before study)
a
Mean values are reported, unless otherwise indicated
b
It was tested whether the number of co-morbid conditions depends on the group membership
* Chi squared test ** Mann–Whitney test #
Fisher’s exact test
are considered clinically normal ([35 mg/dL)] [32] (Table 1). Co-morbidities were reported in 36.8 % of patients (41 % in Bosnia and 26 % in Germany). The most common co-morbidities were hypothyroidism in Bosnia (patients were stabilized on L-Thyroxine 25–50 mcg before the study), and asthma, hay fever, and lactose intolerance in Germany. There were no significant between-group differences in the incidences of diabetic ketoacidosis and severe hypoglycaemia in the previous 6 months (Table 1).
Pharmaceutical care intervention In total, 38 patients in the intervention group completed C6 face-to-face PhC, 1 patient completed 5 face-to-face PhC visits. In majority of patients study assessments were performed by documenting C6 CRFs per patient and evaluating C6 BG records. At least 6 PCPs were written per patient, containing C6 goals, and C2 PCPs (visits 3 and 6) were submitted to physicians. Five follow-ups per patient were performed by the pharmacists.
123
Change from baseline in HbA1c (%)
0.4281 0.038
0.0001 0.0122
0.4255
0.0030
0.0075 0.00002 0.0153
6 months versus baseline 3 months versus baseline 6 months
6 months
3 months *p = 0.00002
0.6
*p = 0.0075
0.4 0.2 0 -0.2 -0.4 -0.6 -0.8
0.2450
Control n = 28
Intervention n = 39
Control n = 26
Primary outcome Relative to usual care, the PhC intervention reduced HbA1c. Mean HbA1c decreased from 9.4 % at baseline to 8.9 % at 6 months in the intervention group, compared with an increase from 9.4 to 9.9 % in the control group (change from baseline -0.54 vs. ?0.32 %, p = 0.0075) (Table 2; Fig. 2). The between-group-difference was greatest after 3 months (-1.09 vs. ?0.23 %, p = 0.00002); (Table 2; Fig. 2). Mean HbA1c reductions were numerically (but not statistically) greater in poorly controlled patients (baseline HbA1c [9 %) than in moderately controlled patients (baseline HbA1c C7.5 to \9 %) after 3 (p = 0.050) and 6 months (p = 0.6978). When analysed by country, a significantly greater mean HbA1c improvement in the intervention group versus the control group was observed at both 3 and 6 months in Bosnia-Herzegovina, but only at 3 months in Germany (all p \ 0.04) (Table 2). Because HbA1c outcomes differed between countries, (potentially influencing the overall HbA1c outcome), linear regression was performed to adjust for the potential country effect, showing that the factor ‘‘country’’ had no effect on the differences in HbA1c between baseline and 6 months (p = 0.5879); only the factor ‘‘group’’ had a significant influence (p = 0.0078). Severe hypoglycemic events ** Mann–Whitney test
HbA1c values between intervention and control at different time points: baseline, after 3 months and after 6 months
Difference in HbA1c values: intervention versus control after 3 months relative to baseline and after 6 months relative to baseline
Intervention n = 40
Fig. 2 HbA1c % difference in the intervention and control groups after 3 months versus baseline and after 6 months versus baseline
c
Data are mean values, unless otherwise indicated a
n, number of patients; HbA1c, Glycosylated haemoglobin
9.1 (7.1–12.5)
9.9 (7.7–12.9) 8.9 (7.2–11.4) 9.5 (7.5–12.2) Germany (range)
9.3 (7.5–14.7) BosniaHerzegovina (range)
By country
8.8 (6.6–13.1)
8.9 (6.6–13.1) Total HbA1c (%) (range)
b
0.5160 9.7 (6.8–11.7)
0.4178 10.0 (7.7–13.0) 9.6 (7.2–12.8) 8.0 (5.7–11.2) 9.3 (7.5–13.6)
9.4 (7.5–14.7) 8.3 (5.7–11.4) 9.4 (7.5–13.6)
123
*Mann Whitney Test
9.8 (7.8–12.8)
0.0037
0.0020 0.5833 9.9 (6.8–13.0) 9.7 (7.2.–12.8)
Baseline Baseline (n = 28) 3 months (n = 40) Baseline (n = 40)
6 months (n = 39)
Control groupa Intervention groupa
3 months (n = 28)
6 months (n = 26)
p value**c
3 months
-1.0
Parameters
Table 2 HbA1c flow and HbA1c difference after 3 and 6 months in the Intervention and Control Groups in the Total Population and by Country
p value** of HbA1c differences between groupsa,b
Int J Clin Pharm
There were no significant between-group differences in the number of severe hypoglycemic events during the study (p = 0.1276) or in the previous 6 months (p = 1). During the study, one severe hypoglycemic event was reported by 7 of 40 patients in the intervention group and 1 of 28 patients in the control group. In the previous 6 months, one
Int J Clin Pharm
event was reported in 6 and 4 patients in the respective groups, while two events were reported in one patient in each group. Well-being Index (WHO-5) (intervention group) In the intervention group, baseline mean well-being score was 52.8 % (range 16–84 %); however, 45 % of patients scored \50 % (indicating low mood) and 13 % scored \28 % (indicating likely depression) [30]. Wellbeing scores improved significantly to 59.2 % (range 24–100 %) at 3 months (p = 0.0002) and to 63.3 % (range 24–96 %) at 6 months (p = 0.00002). At 3 months, 33 and 3 % of patients scored \50 and \28 %, respectively; the corresponding percentages at 6 months were 26 and 3 %. Moreover, the change from baseline in well-being at 6 months was [10 %, which is regarded as clinically relevant [30]. Linear regression revealed that the covariates country (p = 0.9563), age (p = 0.7033), insulin device (p = 0.7169), duration of T1DM (p = 0.9850), and hypoglycaemia at baseline (p = 0.2337) had no influence on the wellbeing scores after 6 months.
Discussion This study demonstrates that PhC in adolescents with T1DM and poor glycemic control contributed glycemic control, as shown by significant HbA1c reductions with no increase in severe hypoglycaemia. Our findings however, in particular the HbA1c reduction, are in contrast to previous research [15, 16]. The lack of benefit in these studies may result from the absence of a structured PhC [17, 18]. In our study, pharmacists assessed clinical data and developed PCPs with problem-solving interventions in concert with the patients to define responsibilities [18] and improve adherence. The PCPs were discussed and progress evaluated with physicians. Therefore, implementation of an appropriate, structured, and multidisciplinary PhC [17, 18], along with patient empowerment, seems crucial for the improved glycemic control we observed. Large studies (n = 289 and 288) in Australia and Belgium with [ 50 community pharmacies evaluating PhC impact on clinical outcomes in adults with T2DM also reported HbA1c reductions (ranging 0.5–1.0 %) [33, 34]. Thus, our trial further supports the value of pharmacists in a multidisciplinary team to reduce disease burden in T1DM. The well-being perception in our study improved by [10 % after 6 months, which is considered to be clinically relevant [30]. However, because well-being was evaluated only in the intervention group, we sought to compare our results with those from other studies as a reference; two studies that included pharmacist
intervention [35, 36] showed similar results to ours. A study by Petkova et al. [35] reported an improvement of 11.7 % in patients with T1DM, and a study by Fresenius [36] reported an improvement of [10 % in patients with T1DM and T2DM. These findings, which are comparable with our results, indicate that PhC is likely to have a positive influence on patient well-being, as the wellbeing in patients who received clinical care from pharmacists did not worsen over the course of the studies. The simple coin-toss randomization method was selected in our study for practical, organizational and coordination reasons, as the same method needed to be used in both study centers. Given the small sample size, this simple randomization resulted in unequal numbers of patients in the study groups. However, unequal randomization ratios significantly reduce study power only if the ratio is 3:1 or more [37–39], which is not expected in our study with an apparent ratio of 4:3. Most importantly, our simple randomization resulted in well balanced study groups (Table 1). Remarkably, in both countries, a greater HbA1c reduction was achieved after 3 rather than 6 months, possibly because of decreasing patient motivation. This finding supports previous studies [6, 16, 40] showing adherence challenges in adolescents (intrinsic to the developmental stage [6]) and requires a longer follow-up with new approaches to sustain the initial improvement. Although HbA1c reduction was achieved in both countries, the effect was more prominent in Bosnia-Herzegovina, which is probably due to the clinical pharmacist’s on-site presence, leading to more frequent and a more intensive patient-pharmacist-physician contact. As a matter of fact, more patients in Bosnia received additional PhC visits and calls then in Germany and as is known, more frequent visits are associated with glycemic control [41–43]. Also, in Bosnia, patients start from a lower baseline of care, and any patient-focused approach is likely to yield a substantial improvement. Limited health care and overall resource shortage may have contributed to better acceptance and adherence by the youth in Bosnia. Acute and long-term T1DM complications result in considerable premature mortality [9, 44]. The Epidemiology of Diabetes Interventions and Complications (EDIC) study confirmed that lowering HbA1c by 1.7 % (HbA1c difference 8.1 vs. 9.8 %) decreases the risk of retinopathy by 53 %, microalbuminuria by 54 %, and neuropathy by 60 % [44]. Therefore, the HbA1c reduction of 0.54 % in our study has potential clinical relevance, particularly in high-risk, poorly controlled patients, who may benefit from the PhC, more frequent visits [41–43] and extended support by the pharmacists. Limitations There are limitations to our study. A major limitation is the heterogeneous study settings between the two countries—
123
Int J Clin Pharm
single clinical pharmacist in Bosnia combined with community pharmacists in Germany which limits study’s external and internal validity. Furthermore, a priori sample size calculation was not performed. All available patients in the clinical setting were screened and incorporated. A post hoc power analysis, however, was performed indicating a study power of 81.90 %. The study results were analyzed per protocol and not by intention to treat. A simple coin-toss randomization method resulted in unequal study groups, as the participant numbers were too small to achieve a relatively even split between comparator groups. A larger number of patients investigated in the same practice environment may be associated with a higher validity than the setting in the DIADEMA trial. The short duration implies that the results may become less favorable with a longer study duration, due to decreasing patient motivation [6, 40], warranting further research on methods to sustain the PhC improvements. Due to time constraints, secondary parameters such as well-being was recorded only for the intervention group. Finally, two study sites in only two countries are unlikely to represent the general adolescent T1DM population; healthcare differences (including treatment and counselling) may pose barriers to PhC implementation in other settings [45].
Conclusion Our study suggests that multidisciplinary PhC may add value in the management of T1DM in adolescents and inadequate glycemic control, as shown by HbA1c reduction. Integration of pharmacists into multidisciplinary T1DM teams may potentially help reduce the rate of complications of diabetes and improve existing care structures across multiple healthcare systems. However, the optimal methods including homogenous settings, a lager cohort and a longer follow-up, on how to achieve sustained, long-term improvements require further study. Acknowledgments The authors thank the patients. They also thank the participating pharmacists (in alphabetical order): Henrich Dieter Backes, Zora Bahser, Hans-Dieter Brink, Dieter Conze, Kathrin Furth, Klaus Dieter von Laguna, Anja Mu¨ller, Norbert Mu¨ller, Anne Rhein, Mara Scholz, Friederike Sieben, Ralf Weckop, Katja Weissenborn, and the pediatric clinic directors: Prof. Dr. med. Tim Niehues and Prof. Dr. med. Senka Mesihovic-Dinarevic. We are indebted to K. A. Lyseng-Williamson and L.Yang for editorial assistance. Funding Emina Obarcanin was financially supported by the Deutsche Akademische Auslandsdienst (DAAD), the Lesmu¨ller Stiftung, and the Heinrich-Heine University, Du¨sseldorf, Germany. The study was partially supported by the Lesmu¨ller Stiftung. Conflicts of interest
123
The authors have no conflicts of interest.
References 1. The DIAMOD project group. Incidence and trends of childhood Type 1 diabetes worldwide 1990–1999. Diabet Med. 2006;23:857–66. 2. Patterson C, Dahlquist G, Gyu¨ru¨s E, Green A, Solte´sz G. EURODIAB Study Group. Incidence trends for childhood type 1 diabetes in Europe during 1989–2003 and predicted new cases 2005–20: a multicentre prospective registration study. Lancet. 2009;373:2027–33. 3. Ehehalt S, Dietz K, Willasch AM, Neu A, Baden-Wu¨rttemberg Diabetes Incidence Registry (DIARY) Group. Epidemiological perspectives on type 1 diabetes in childhood and adolescence in Germany. Diabetes Care. 2010;33:338–40. 4. Tahirovic H, Toromanovic A. Incidence of type 1 diabetes mellitus in children in Tuzla Canton between 1995 and 2004. Eur J Pediatr. 2007;166:491–2. 5. Daneman D, Hamilton J. Is poor metabolic control inevitable in adolescents with type 1 diabetes? An Pediatr. 2001;1:40–4. 6. Borus J, Laffel L. Adherence challenges in the management of type 1 diabetes in adolescents: prevention and intervention. Curr Opin Pediatr. 2010;22:405–11. 7. Patino AM, Sanchez J, Eidson M, Delamater AM. Health beliefs and regimen adherence in minority adolescents with type 1 diabetes. J Pediatr Psychol. 2005;30:503–12. 8. Schober E, Wagner G, Berger G, Gerber D, Mengl M, Sonnenstatter S, et al. Prevalence of intentional under-and overdosing of insulin in children and adolescents with type 1 diabetes. Pediatr Diabetes. 2011;12:627. 9. Harjutsalo V, Forsblom C, Groop P. Time trends in mortality in patients with type 1 diabetes: nationwide population based cohort study. BMJ. 2011;343:5364. 10. Petersen M. Economic costs of diabetes in the U.S. in 2012. Diabetes Care. 2013;36:1033–46. 11. Daneman D. Type 1 diabetes. Lancet. 2006;367:847–58. 12. Wermeille J, Bennie M, Brown I, McKnight J. Pharmaceutical care model for patients with type 2 diabetes: integration of the community pharmacist into the diabetes team-a pilot study. Pharm World Sci. 2004;26:18–25. 13. Chisholm-Burns MA, Kim Lee J, Spivey CA, Slack M, Herrier RN, Hall-Lipsy E, et al. US pharmacists’ effect as team members on patient care: systematic review and meta-analyses. Med Care. 2010;48:923–33. 14. Machado M, Bajcar J, Guzzo GC, Einarson TR. Sensitivity of patient outcomes to pharmacist interventions. Part I: systematic review and meta-analysis in diabetes management. Ann Pharmacother. 2007;41:1569–82. 15. Gay CL, Chapuis F, Bendelac N, Tixier F, Treppoz S, Nicolino M. Reinforced follow-up for children and adolescents with type 1 diabetes and inadequate glycaemic control: a randomized controlled trial intervention via the local pharmacist and telecare. Diabetes Metab. 2006;32:159–65. 16. Sims LM, Haines SL. Challenges of a pharmacist-directed peer support program among adolescents with diabetes. J Am Pharm Assoc. 2011;51:766–76. 17. Cipolle RJ, Strand LM, Morley PC. Pharmaceutical care practice: The clinician’s guide. 2nd ed. New York: McGraw-Hill; 2004. 18. van Mil JW, Schulz M, Tromp TF. Pharmaceutical care, European developments in concepts, implementation, teaching, and research: a review. Pharm World Sci. 2004;26(6):303–11. 19. Bosnia-Herzegovina Agency for Drugs and Medical Devices. Clinical trials [Agencija za lijekove i medicinska sredstva Bosne i Hercegovine]. http://www.almbih.gov.ba/klinicka-ispitivanja. Accessed 19 Sep 2014.
Int J Clin Pharm 20. Bosnia-Herzegovina Agency for Drugsand Medical Devices. Clinical trial and medical devices directive of Bosnia and Herzegovina. Official Gazette of Bosnia and Herzegovina No. 4/10 [Sluzbeni Glasnik Bosne i Hercegovine Nr. 4/10 http:// www.almbih.gov.ba/_doc/regulative/pravilnik_klinicka_bos.pdf. Accessed 19 Sep 2014. 21. Ineternational Society of Pediatric Diabetes. Clinical practice consensus guidelines 2009. http://www.ispad.org/content/ispadclinical-practice-consensus-guidelines-2009. Accessed 19 Sep 2014. 22. Faul F, Erdfelder E, Lang A-G, Buchner A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39:175–91. 23. Blume J, Peipert JF. Randomization in controlled clinical trials: why the flip of a coin is so important. J Am Assoc Gynecol Laparosc. 2004;11:320–5. 24. Berger W. A review of methods for ensuring the comparability of comparison groups in randomized clinical trials. Rev Recent Clin Trials. 2006;1:81–6. 25. Clark M, Westerberg B. How random is the toss of a coin? CMAJ. 2009;181:306–8. 26. Viera A, Bangdiwala I. Eliminating bias in randomized controlled trials: Importance of allocation concealment and masking. Fam Med. 2007;39:132–7. 27. German National Treatment Guideline Program. Diabetes structured education program Version 3 Jun 2013 [Program fuer nationale Versorgungslinie, Diabetes strukturierteSchulungsprogramme].http:// www.versorgungsleitlinien.de/themen/diabetes2/dm2_schulung/ pdf/nvl-t2d-schulung-lang-3.pdf. Accessed 19 Sep 2014. 28. Workgroup on Hypoglycemia, American Diabetes Association. Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia. Diabetes Care. 2005;28(5):1245–9. 29. De Wit M, Pouwer F, Gemke RJ, Delemarre-van de Waal HA, Snoek FJ. Validation of the WHO-5 Well-Being Index in adolescents with type 1 diabetes. Diabetes Care. 2007;30:2003–6. 30. WHO (Five) Well-Being Index (1998 version). Hillerod: Psychiatric Research Unit, WHO Collaborating Center for Mental Health, Frederiksborg General Hospital. 1998. http://www.psy kiatri-regionh.dk/NR/rdonlyres/ACF049D6-C94D-49B2-B34C96A7C5DA463B/0/WHO5_English.pdf. Accessed 19 Sep 2014. 31. R Foundation for Statistical Computing 2014. A Language and Environment for Statistical Computing. R Core Team Vienna, Austria: http://www.r-project.org. Accessed 19 Sep 2014.
32. American Diabetes Association. Management of dyslipidemia in children and adolescents with diabetes. Diabetes Care. 2003;26:2194–7. 33. Krass I, Armour CL, Mitchell B, Hughes J, Peterson G, Song YJ, et al. The Pharmacy Diabetes Care Program: assessment of a community pharmacy diabetes service model in Australia. Diabetic Med. 2007;24:677–83. 34. Mehuys E, Van Bortel L, De Bolle L, Van Tongelen I, Annemans L, Remon JP. Effectiveness of a community pharmacist intervention in diabetes care: A randomized controlled trial. J Clin Pharm Ther. 2011;36:602–13. 35. Petkova V, Ivanova A, Petrova G. Education of patients with diabetes in the community pharmacies (pilot project in Bulgaria). Ankara Ecz Fak Derg. 2006;35:111–24. 36. Fresenius K. Impact of pharmaceutical care on clinical and therapy outcomes in patients with diabetic foot syndrom (DFS) [Dissertation]. Institute for Chemistry, Pharmacy and Geo-Science, University of Mainz, Germany: Fresenius K, 2007. 37. Pocock SJ. Clinical trials: a practical approach. Chichester: Wiley; 1995. 38. Gail M, Williams R, Byar DP, Brown C. How many controls? J Chronic Dis. 1976;29:723–31. 39. Dumville JC, Hahn S, Miles JN, Torgerson DJ. The use of unequal randomisation ratios in clinical trials: a review. Contemp Clin Trials. 2006;27:1–12. 40. Frank M. Psychological issues in the care of children and adolescents with type 1 diabetes. Paediatr Child Health. 2005;10:18–20. 41. Phan TL, Hossain J, Lawless S, Werk NL. Quarterly visits with glycated hemoglobin monitoring: The sweet spot for glycemic control in youth with type 1 diabetes. Diabetes Care. 2014;37:341–5. 42. Kaufman FR, Halvorson M, Carpenter S. Association between diabetes control and visits to a multidisciplinary pediatric diabetes clinic. Pediatrics. 1999;103:948–51. 43. Svoren BM, Butler D, Levine BS, Anderson BJ, Laffel LM. Reducing acute adverse outcomes in youths with type 1 diabetes: A randomized, controlled trial. Pediatrics. 2003;112:914–22. 44. The Epidemiology of Diabetes Interventions and Complications (EDIC) study. Sustained effect of intensive treatment of type 1 diabetes mellitus on development and progression of diabetic nephropathy. JAMA 2003; 2159–2167. 45. van Mil JW, de Boer W, Tromp TH. European barriers to the implementation of pharmaceutical care. International Journal of Pharmacy Practice. 2001;9:163–8.
123
