Management of Subjects with Type 2 Diabetes Hospitalized in Internal Medicine Units: A Cluster-Randomized, Multicenter Study Before &amp; After an Educational Program

doi:10.21203/rs.3.rs-4842971/v1

Introduction: Few data are available relatively to the adherence to treatment guidelines in individuals with type 2 diabetes mellitus (T2DM) admitted to Internal Medicine Wards (IMW) while no information is available relatively to the possible efficacy of an educational intervention aimed at implementing adherence in this setting.

Methods: To explore guidelines adherence and the relative impact on glycemic control in subjects with T2DM hospitalized in IMW before and after an educational intervention, we conducted a 3-phase, cluster-randomized, multicenter study. During phase 1, we retrospectively collected data from patients with T2DM hospitalized for any cause in IMW for ≥ 5 days. In phase 2, an educational training, based on the method of the educational outreach visits (EOV), was developed in 36 out of the 54 centers involved. In phase 3, done 6 months after the training, we replicated the collection of data performed in phase 1.

Results: Overall, we analyzed data from 1909 and 1662 individuals with T2DM during the phase 1 and phase 3 of the study, respectively. No changes were observed in the difference between mean fasting glycemia levels at discharge vs at admission at Phase 3 comparing EOV vs NO EOV groups. A statistically significant increase in adherence to guidelines was observed from Phase 1 to Phase 3 and a trend of higher adherence was detected when comparing the EOV and the no EOV groups. A structured educational intervention improves adherence to guidelines for managing T2DM in individuals admitted to IMW but has no effect on short-term glycemic control.

GLP-1RA

glycemic control

guidelines

SGLT-2i

treatment adherence

type 2 diabetes

Type 2 Diabetes mellitus (T2DM) is a world pandemic increasing the risk of a range of complications, especially cardiovascular. Several studies have shown that early and aggressive control of all the most relevant risk factors, in particular hyperglycemia, dyslipidemia and hypertension can effectively prevent the development of these complications [1-3].

Obtaining optimal glycemic control, stable over time, is still a difficult result to achieve. In recent years, the availability of new safe and effective drugs has significantly changed the therapeutic approach to the disease. Patients’ phenotyping, according to a range of characteristics and comorbidities, is now the recommended standard to choose the most suitable therapy [4-8].

The results of selected studies on recently introduced classes of glucose-lowering drugs, i.e. GLP1-RA and SGLT-2i, have contributed to a further development of this strategy, with positive findings in terms of reduction of the cardiovascular risk, of hospitalizations for heart failure, and of kidney damage [4-6]. These data, together with the effects of such drugs on the risk of hypoglycaemia or in promoting weight reduction have led international Scientific Societies to produce guidelines [7-8] that have substantially changed the paradigms of the treatment of hyperglycaemia in T2DM. These guidelines recommend the opportunity to privilege, for the therapeutic choice, those classes of drugs that have demonstrated a particular effect in the specific condition that characterizes the subject with T2DM (for example, a history of previous cardiovascular event, or the presence of heart failure, or of obesity).

Despite the strong evidence available and the indications of the Scientific Societies, the application of these recommendations is far from being optimal [9]. Specifically, the recent changes in the therapeutic scenario appear particularly relevant for some care settings. Among them, the area of Internal Medicine in which the complexity of the patients as well as some aspects of healthcare organization (e.g., the limited duration of hospital admissions, the transition from hospital to primary care and/or outpatient clinics) make the choice and appropriateness of treatments a challenging issue. However, in this setting, the adherence to existing guidelines on the management of patients with T2DM and the possible impact of an educational intervention have been poorly explored.

To address this issue, we conducted a multicenter study to take a real-life snapshot of the management of patients with T2DM hospitalized for any cause in Italian Internal Medicine Wards (IMW) and evaluated possible deviations between the current clinical practice and the recommended guidelines (i.e. ESC-EASD and EASD-ADA) in force at the moment of study start [5-8]. Moreover, we evaluated the effects of a specific educational program for health care personnel in IMW, aimed at improving the management of these patients and the adherence to guidelines.

The MINDER study (Management of patIeNts with type 2 Diabetes mEllitus hospitalized in InteRnal Medicine Units) was designed and developed by the Italian Federation of Internal Medicine Hospital Clinicans’Associations (FADOI). The MINDER study was performed in 50 Italian Internal Medicine Units and carried out in the period September 2020- March 2023. Centers were selected according to their ability to prescribe all classes of antidiabetic drugs.

The study was conducted according to the provisions of the Helsinki Declaration. Approval was obtained from the Ethics Committees of each participating center.

It was designed as a replicate of two cross-sectional surveys interspersed with an educational program, a model successfully applied by FADOI to a number of clinical conditions [10,11,12,13].

As shown in Figure S1, the study was composed by three steps. Phase 1 concerned a retrospective data collection pertaining to patients ≥ 18 years old with diagnosis of T2DM hospitalized for any cause in IMW for ≥ 5 days before February 2020. Given that data collection was retrospective, patients were managed according to routine clinical practice. As per study protocol, patients in treatment with insulin or patients treated with rapid acting insulin during the hospitalization or patients hospitalized for SARS-CoV-2 infection were excluded.

In phase 2 an educational training, defined by the study Steering Committee, and focused on current Diabetes Guidelines recommendations and on possible deviations from the best clinical practice observed during phase 1, was conducted in 36 out of the 54 participating Centers (cluster randomization, performed at the beginning of the study). The ratio 2:1 was selected in order to offer to the majority of Centers the opportunity to undergo a program of training and therefore maximize the potential positive effect of the educational program on patient management. However, at the end of the study, the educational program was offered also to the remaining 18 Centers. The program was based on the method of the educational outreach visits (EOV), a 3-hour face-to-face meeting between a trained diabetes specialist from outside the Center and the staff of the Center itself. This method was selected being considered as one of the most effective to modify professional practice and improve health care outcomes [14]. In addition, all members of the staff of the Center received a distance learning program, to reinforce and deepen the contents of the outreach visit. Prior to the outreach visits, each diabetes specialist in charge of training (n = 9) was asked to attend a 1-day briefing workshop to gain experience with the educational program and standardize the contents to be delivered.

Phase 3 occurred around 6 months after the training and concerned a new data collection replicating that performed in phase 1. In both phase 1 and phase 3, the data collection was based on the review of the medical records of the last 40 consecutive patients (in the period December 2018 – December 2020 for Phase1 and February 2022 - February 2023 for Phase 3) complying with study inclusion criteria and hospitalized in each Center.

In order to minimize the “awareness bias” (i.e. the possibility that participation to the project significantly influenced clinical attitudes in phase 3), only one physician of the Centers of the “Control” group (not receiving EOV and distance learning) was made aware of the study design and procedures, and he/she was eventually supported in the study data collection by nurses or by physicians who did not have the possibility of prescribing the new categories of antidiabetic drugs.

In details, collected information included: gender, age, anthropometric parameters, reason for hospitalization, concomitant diseases and relevant drug therapies, therapies for T2DM at admission to hospital and at discharge, laboratory results (fasting glycemia [FG] and glomerular filtration rate at hospital admission and at discharge, glycated hemoglobin) occurrence of hypoglycemia, duration and outcome of hospitalization. To improve quality of data collection, a study-specific electronic case report form (e-CRF) with central remote control was used.

For each case, an Independent Committee of Experts assessed the adherence to guidelines of antidiabetic therapy at discharge from the hospital. In particular, two Experts, unaware of the study group (receiving or not receiving the educational program), examined each case independently. In case of discordant evaluation, the case report was examined by a third Expert. This method and the relevant specific procedures (i.e. clinical information to be analyzed, criteria for adjudication) were considered reliable and selected by the Experts prior to the beginning of the study, following an inter-rater reliability test (by Fleiss' kappa score) performed to measure the level of agreement between multiple judges for the same item (9). In particular, each Expert was asked to examine 10 clinical cases and the level of agreement was measured by using Fleiss' kappa statistics. The result achieved with this assessment (repeated two times after two cycles of training) was a moderate agreement (k< 0,5), therefore we considered that each clinical case could not be evaluated by only one Expert.

2.1 SAMPLE SIZE and Statistical Analysis

Sample sizes of 1440 in Test Arm (EOV GROUP) and 720 in Control Arm (NO-EOV GROUP), which were obtained by sampling 36 centers (clusters) with an average of 40 subjects each in EOV GROUP and 18 centers with an average of 40 subjects each in NOT EOV GROUP, achieve about 80% power to detect a difference between the group means of at least 10. This estimate was calculated assuming, in phase 3, a mean difference between the first and the last measurement of fasting glycemia of around 17 mg/dL in patients hospitalized in NO-EOV GROUP Centers, and of around 27 mg/dL in patients hospitalized in EOV GROUP Centers. This estimate has been done according to previous data collected in a study performed some years ago in a very similar setting [9]. The standard deviation of subjects is about 75. The intra-cluster correlation coefficient is 0,001 (Ballpark estimate). The coefficient of variation of cluster sizes is 0,650 (Ballpark estimate). A two-sided t-test was used with a significance level of 0,050. This test used degrees of freedom based on the number of clusters. The sample size calculations were performed using the commercial software PASS 14. According to the study design, to have two mirroring phases, in the phase 1 the sample size will be of 2160 subjects, as well as in the phase 3. All test will be two-sided and performed at the significance level of α = 0.05

For primary efficacy endpoint, variation in fasting glycemia during hospitalization between the two groups of Centers (those receiving and not receiving the educational program) during phase 3 of the study, a mixed linear model with identity link function, gaussian distribution and parameterized according to a cluster-randomized design was estimated with “treatment indicator” as fixed effect, “center indicator”, (namely the cluster randomization unit), as random effect and the dependent variable computed for each patient (subject within cluster) using the change between the first and the last measurement of fasting glycemia. Point estimate of treatment difference with associated two-sided 95% Confidence Interval was reported as least-squares estimates and computed with degrees of freedom adjusted according to Kenward-Roger method in order to take into account the cluster nature of the study. The procedure GLIMMIX of SAS Software was employed for the aforementioned analyses and computations.

Chi-square test for categorical parameters, and T-test or Wilcoxon rank sum test, for continuous ones, were used to compare baseline characteristic and outcomes between phases and groups (phase 3 only).

The difference between variables at discharge vs. admission among the same subjects for phase 1, phase 3 EOV and phase 3 NO EOV was explored applying paired t-test (continuous) and McNemar test (categorical).

To outline change in the prescription’s treatment percentage changes were calculated as

To study the adherence to guidelines of antidiabetic therapy at discharge, a multilevel logistic regression was applied with the following variables as covariates: age (75 vs <75 years-old), presence of concomitant disease ((3 vs <3), obesity (BMI>30), previous cardiovascular event, duration of hospitalization (7 vs <7 days), FG at admission categorized as <70mg/dl (reference), 70-125 mg/dL, >125 mg/dL.

The demographic and clinical-pathological characteristics were appropriately summarized by means of descriptive statistics. Mean, median and related distribution parameters were reported for continuous variables, while frequencies and percentages were calculated for the different categorical variables.

In Phase 1 and in Phase 3 of the study data were collected from 1909 and 1662 patients with T2DM, respectively (total number 3571). General characteristics of patients at baseline and at discharge are described in Table 1.

As confirmation of the complexity of patients with diabetes hospitalized in IMW, the study population had a mean age of 76 years, suffered a median of three chronic diseases, and around 40% of patients had a previous cardiovascular event.

In both Phase 1 and in Phase 3, mean FG at the time of hospitalization was higher than that at discharge, as a result of the clinical management during the hospitalization in IMW (Figure 1).

We detected no significant differences for the primary endpoint of the study, i.e. the variation in FG during hospitalization between the two groups of Centers (EOV vs NO EOV) during Phase 3 of the study. The mean variation was 25,46 mg/dL in EOV group and 26,19 mg/dL in NO EOV group (p=0,93) (Point estimate of the difference = 0,73; 95% CI: -17,9;19,4) (Figure 1).

Table 2 shows the classes of medication for the treatment of T2DM at the admission and at discharge in Phase 1 and in Phase 3 for each group of Centers. In the EOV group a statistically significant higher percentage of patients received specific therapy for T2DM.

The percentage of patients adherent to evidence-based T2DM therapy at discharge are summarized in Figure 2. A statistically significant increase of adherence was observed from Phase1 to Phase 3 and a trend of higher adherence was detected in the EOV group VS NO EOV.

Table 3 shows the results of univariate analyses and a multivariable logistic analysis aimed at evaluating potential predictors of non-adherence to optimal T2DM therapy in the Phase 3 of the study. Previous cardiovascular events and concomitant diseases (≥3) were predictors of non-adherence in Phase1 (data not shown), but not in Phase 3 (Table 3). In Phase 3, age (≥75 years) and low FG at admission (< 70 mg/dL) were associated with a significant difficulty in achieving guideline-oriented management of T2DM. On the other hand, obese patients seem to receive a more appropriate T2DM treatment than patient with lower BMI.

Following the educational outreach visit, participating IMW staff (both physicians and nurses) were asked to anonymously rate an overall assessment of the event based on appropriateness of the training modality and relevant contents, clarity of exposition by the tutor, and level of interaction. The mean rate of the overall level of satisfaction with the training received and reported by the participating healthcare professional was 8,4 (from 1 = unwelcome to 10 = welcome). In Figure S2, a qualitative stratification of the level of satisfaction for each of the four aforementioned items is shown.

In this 3-step study, we showed that a structured educational intervention is able to improve adherence to existing guidelines for T2DM management in patients hospitalized in IMW. However, possibly due to other educational activities on these issues occurring during the study period, even those who did not receive the educational intervention still improved their adherence, though to a lesser extent than the interventional group.

Glycemic control improved over time, both at admission and discharge. Relatively to the main efficacy endpoint of the study, we did not detect relevant differences in glycemic control, neither between study phases nor between educated vs non-educated groups considering the ∆ admission vs discharge. Furthermore, the NO EOV group had a larger difference than that hypothesized during study design. This latter result can be attributed to an adequately established and appropriate clinical behavior for glycemic control during hospitalization. Furthermore, improvement in the management of these patients may be ascribed to the participation of clinicians in other types of educational training interventions during the period of the study. Furthermore, we must acknowledge that we did not enroll the proposed number of individuals, largely due to the impact on hospital activities driven by the Covid-19 pandemic [15], an aspect that could have affected our results.

We observed a higher use of novel glucose-lowering drugs with established cardioprotective properties after the educational program. Although this did not provide a short-term benefit on glycemic control, it is likely that the effects of increased adherence to the guidelines might provide medium or long-term benefit in terms of reduction of hard cardiovascular endpoints, which were not explored in this study. Prevalent cardiovascular events and concomitant diseases (≥3) were predictors of non-adherence in Phase 1, but not in Phase 3, likely due to the effect of the educational program. Similarly, we observed a consistent increase of the percentage of referrals to divisional ambulatory and to regional diabetes specialists in the group exposed to the educational program, suggesting an improved awareness relatively to this aspect. Other predictors of non-adherence in phase 3 were a low blood glucose level at admission and older age. Relatively to the latter variable, possible influencing factors may be the range of comorbidity, overall health status, number of drugs used, and complexity of the glucose-lowering regimen, which are more common in the elderlies [17].

A key aspect to consider when developing an educational program is the satisfaction of the professionals involved. The 4-item survey conducted among the staff of the IMW evidenced a high rate of satisfaction relatively to the appropriateness of the training modality, the clarity of exposition by the tutor, and level of interaction. This should reassure about the ability of the EOV approach to effectively deliver the necessary recommendation to the staff. However, given our study design, we cannot establish whether alternative methods of dissemination of guidelines, either active or passive, would have obtained similar results [18].

Beyond the difficulties encountered due to the Covid-19 pandemic and the limited sample size, which was powered to detect a difference in glycemic control but eventually not an increase in treatment adherence, this study presents additional, intrinsic limitations related to its design. First, we cannot rule out that doctors’ awareness of participating in this study might have made them more attentive to the management of diabetes, thus affecting results achieved in terms of compliance with recommendations (awareness bias). This concern would apply particularly to the results obtained during the phase 3. However, cluster randomization of centers, receiving or not an educational program, should reassure about the reliability of the results relative to the effect of training. Second, retrospective data collection from hospital clinical records might have missed some information required by the study protocol. On the other hand, such an approach should have limited the effect of the awareness bias, providing a reliable snapshot of real-life clinical practice. Another possible limitation is that only patients not on insulin therapy were enrolled in the study, limiting the generalization of findings. However, this choice was made with the purpose to select patients who would more plausibly benefit from an improved adherence to the guidelines. More generally, the number and the geographic distribution of the IMW involved reassure about the overall representativeness of included individuals compared with the general population with T2DM. Finally, the two study groups, those exposed to the educational intervention and the control group, differed in some baseline characteristics. Unfortunately, this risk can occur with the application of the cluster randomization. This might have modestly affected the results relative to metabolic control, but they unlikely have influenced adherence to guidelines, as documented by the predictors of adherence emerged in the phase 3 of the study.

In summary, our results suggest that implementing a structured educational intervention effectively enhances adherence to established guidelines for managing T2DM in individuals admitted to the IMW, even though it does not improve short-term glycemic control. Considering the well-established cardiovascular and renal benefit of novel drugs [19,20], educational interventions aimed at improving adherence to existing guidelines might provide a long-term benefit in terms of hard outcomes for a wide population. Future, larger studies exploring this and other approaches in similar, as well as other, settings should be conducted to establish the usefulness of structured educational interventions.

Acknowledgments

We express our gratitude to AstraZeneca Italy, Takeda, Novo Nordisk Italy and Boehringer Ingelheim Italia S.p.A for supporting the FADOI-MINDER Study with an unrestricted research or educational grant. This research was also supported, in part, by the Italian Ministry of Health through Ricerca Corrente to IRCCS MultiMedica.

Authors’ contribution

Antonio Ceriello: project definition, data analysis, manuscript drafting. Andrea Fontanella: project definition, data collection. Tiziana Marcella Attardo: data collection, tutoring of educational programme. Giampietro Beltramello: tutoring of educational programme. Aldo Fierro: tutoring of educational programme. Francesco Prattichizzo: manuscript drafting. Valeria Pellegrini: manuscript drafting. Maria Serena Fiore: tutoring of educational programme. Ernesto De Menis: tutoring of educational programme. Ada Maffettone: tutoring of educational programme. Luigi Magnani: tutoring of educational programme. Ruggero Pastorelli: tutoring of educational programme. Francesco Ventrella: tutoring of educational programme. Massimo Rondana: data collection. Mauro Maurantonio: data collection. Mauro Scanferlato: data collection. Francesco Finizola: data collection. Maria Antonia Salvia: data collection. Riccardo Nevola: data collection. Giorgia Prampolini: data collection. Marco Laccetti: data collection. Franco Mastroianni: data collection. Fabio Cartabellotta: data collection. Elisa Zagarrì project definition, project and data management, data analysis, manuscript drafting. Francesco Dentali: data collection, critical revision of the manuscript. Dario Manfellotto: data collection, critical revision of the manuscript. All authors read and approved the final submitted version of to the manuscript.

Conflict of interest

All authors declare that they have no conflict of interest concerning the submitted manuscript.

Ethical approval

The study was conducted according to the provisions of the Helsinki Declaration of 1964. Approval was obtained from the Ethics Committees of each participating center.

Data Availability

The datasets generated during and analyzed during this study are available from the corresponding author on reasonable request.

Members of the FADOI – MINDER Study Group (in alphabetical order)

Alessandri Massimo/ Petri Roberta e Fabrizio Valleggi (Ospedale “Sant’Andrea” di Massa Marittima (GR)),

Antonucci Giancarlo, Piredda Sabrina (Ente Ospedaliero “Galliera” di Genova),

Berra Sergio Antonio /Perfetti Luca, Bracale Maria Francesca e Girola Andrea (Ospedale di Garbagnate Milanese (MI))

Burattini Maurizio/Luconi Maria Paola e Nicola Tarquinio (IRCCS Istituto Nazionale di Ricovero e Cura per Anziani INRCA di Osimo (AN)),

Campagna Giuseppe, Giuliani Tommasina (Ospedale “S. Maria Goretti” di Latina),

Cartabellotta Fabio/Pepe Armando (Ospedale “Buccheri-La Ferla” Fatebenefratelli di Palermo),

Cavalleri Maurizio/Bernero Elena/Martino Chiara (P.O. di Chiavari (GE)),

Cerci Sabrina/Leonardi Antonella/Durola Laura (Ospedale di Frascati (RM)),

Corradini Paolo/Sbocchia Daniela (Ospedale "F. Petruccioli" di Pitigliano (GR)),

D’Angelo Francesco/La Spina Tiziana/Dionigi Elena (Ospedale di Cernusco sul Naviglio (MI)),

D’Avino Maria/Caruso Giuseppe (AORN “Cardarelli” di Napoli),

De Angelis Roberta/Romanello Daniele e Sara Rotunno (Ospedale "San Pietro Fatebenefratelli” di Roma),

De Carli Stefano/Spangaro Sabrina (Ospedale “Sant’Antonio” San Daniele del Friuli (UD)),

De Franceschi Teresiano/Saleh Youssef (Ospedale “Santa Maria della Misericordia” di Albenga (SV)),

De Giovanni Raffaella/Vogrig Nadia (AUSLRomagna, Ospedale “Cervesi” di Cattolica, Ospedale di Rimini (RN)),

Dentali Francesco/Attardo Tiziana Marcella (Ospedale “Luni-Confalonieri” di Luino),

Ferrari Giovanni/Asti Micaela, Nunzia Cirillo, Daniela Rebasti, Anna Rita Sorrentino (ASST Pavia, Ospedale di Broni e Stradella (PV)),

Ferraro Maria (ASP Cosenza, PO Acri di Cosenza),

Finizola Francesco/ Rosada Javier (USL TOSCANA NORD OVEST Ospedale “Sant’Antonio Abate” di Fivizzano (MS)),

Fontanella Andrea/Bonaiuto Alessio (Ospedale Buonconsiglio Fatebenefratelli di Napoli),

Gatti Pietro/Barletta Mariangela (P.O. di Brindisi)

Giordano Mauro/Ciarambino Tiziana (P.O. Clinicizzato di Marcianise (CE)),

Giorgi Andrea/Guarini Margherita (Presidio Ospedaliero “San Filippo Neri” di Roma).

Giorgi Pierfranceschi Matteo/Migazzi Matteo (Ospedale Maggiore di Cremona),

Greco Antonio/Gerdano Carmela (IRCCS "Casa Sollievo della Sofferenza" di San Giovanni Rotondo (FG)),

Laccetti Marco/Coronella Concetta (AORN “Cardarelli” di Napoli),

Lombardi Martina (Asl Toscana Nordovest, Ospedale di Cecina, AOU Pisana)

Manfellotto Dario/Ventriglia Mariacarla (Ospedale Isola Tiberina, Gemelli-Isola, Roma).

Manini Maurizio/Lillo Davide (Ospedale “San Giovanni di Dio” di Orbetello (GR))

Mastroianni Franco/ Giovanni Larizza (Ospedale Generale Regionale 'F. Miulli' di Acquaviva nelle Fonti (BA))

Maurantonio Mauro/Castellano Claudia (AOU Ospedale Civile di Baggiovara MO))

Mazzone Antonino/Brendaglia Noemi (Ospedale Civile di Legnano (MI))

Moggi Pignone Alberto/Ramundo Antonello e Bandini Giulia (A.O.U. “Careggi” di Firenze)

Montagnani Andrea/Michele Caselli (Ospedale “Santa Maria della Misericordia” di Grosseto)

Morana Ignazio Maria/Angela Pastore/Maria Giuseppina Pecora (ARNAS “Garibaldi” di Catania))

Muriago Michela/Dal Maso Adriana (Ospedale "Pietro Milani" Noventa Vicentina (VI))

Nevola Riccardo/Pafundi Pia Clara (University of Campania “L. Vanvitelli” di Napoli)

Novati Paola/ Cottarelli Daniela (ASST RHODENSE, Presidio di Passirana di Rho)

Oliva Gabriella/Andreozzi Raffaella (ASL Napoli 1 Centro, Ospedale del Mare di Napoli)

Pietrantonio Filomena/Gabrielli Mariagrazia e Giampiero Marino (ASL Roma 6, Ospedale dei Castelli di Ariccia (RM))

Pirisi Mario/ Gian Paolo Fra (A.O.U. “Maggiore della Carità” di Novara, Università del Piemonte Orientale)

Politi Cecilia/Mancini Concetta (Ospedale “F. Veneziale” di Isernia)

Prampolini Giorgia/Turilli Marzia (Ospedale “Sant’Anna” di Castelnovo ‘Ne Monti (RE))

Renda Maurizio/Imburgia Mariangela e Giuseppe Miceli (Fondazione Istituto “G. Giglio” di Cefalù (PA))

Rondana Massimo/Santoro Fabrizio (ASFO di Spilimbergo (PN))

Salvia Maria Antonia/Martino Rosamaria (P.O. “San Salvatore” di Mistretta (ME))

Scanferlato Mauro/Cirillo Margherita (AULSS 4 Veneto Orientale, Ospedale di Portogruaro (VE)),

Tonizzo Maurizio/ Copat Maria e Zanier Ada (Ospedale Santa Maria degli Angeli" di Pordenone)

Trojani Cristina/Antonini Raffaella e Gabriella Fantini (Ospedale “Ceccarini” di Riccione (RN))

Zoratti Raffaele/Pizzin Cristina (Ospedale di Palmanova, Presidio Ospedaliero Latisana-Palmanova ASUFC - Udine)

Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH, Pedersen O (2003) Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med 348:383–393
Gaede P, Lund-Andersen H, Parving HH, Pedersen O (2008) Effect of a multifactorial intervention on mortality in type 2 diabetes. N Engl J Med 358:580–591
Gæde P, Oellgaard J, Carstensen B, Rossing P, Lund-Andersen H, Parving HH, Pedersen O (2016) Years of life gained by multifactorial intervention in patients with type 2 diabetes mellitus and microalbuminuria: 21 years follow-up on the Steno-2 randomised trial. Diabetologia 59:2298–2307
Standl E, Schnell O, McGuire DK, Ceriello A, Rydén L (2017) Integration of recent evidence into management of patients with atherosclerotic cardiovascular disease and type 2 diabetes. Lancet Diabetes Endocrinol 5:391–402
Cosentino F, Ceriello A, Baeres FMM et al (2018) Addressing cardiovascular risk in type 2 diabetes mellitus: a report from the European Society of Cardiology Cardiovascular Roundtable. Eur Heart J. 10.1093/eurheartj/ehy677[Epub ahead of print]
Schnell O, Standl E, Catrinoiu D et al (2019) Report from the 4th Cardiovascular Outcome Trial (CVOT) Summit of the Diabetes & Cardiovascular Disease (D&CVD) EASD Study Group. Cardiovasc Diabetol 18:30
Davies MJ, D'Alessio DA, Fradkin J et al (2018) Management of Hyperglycemia in Type 2 Diabetes, 2018. A Consensus Report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 41:2669–2701
Arnett DK, Blumenthal RS, Albert MA et al (2019) 2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. ; pii: S0735-1097(19)33877-X. 10.1016/j.jacc.2019.03.010. [Epub ahead of print]
Gabler M, Picker N, Geier S et al (2021) Guideline Adherence and Associated Outcomes in the Treatment of Type 2 Diabetes Mellitus Patients With an Incident Cardiovascular Comorbidity: An Analysis Based on a Large German Claims Dataset. Diabetes Ther. ; (4):1209–1226
Gulli G, Frasson S, Borzi V et al (2014) Effectiveness of an educational intervention on the management of type 2 diabetic patients hospitalized in Internal Medicine: results from the FADOI-DIAMOND study. Acta Diabetol 137:137–148
Brancato D, Biondi B, Attardo TM et al (2022) Management of Hypothyroidism in Internal Medicine: Patient Profile and Effects of an Educational Programme in the Cluster- Randomized FADOI TIAMO Study. Front. Endocrinol 13:839300
Civardi G, Zucco F, Valerio A et al (2015) Pain management in internal medicine and effects of a standardised educational intervention: the FADOI-DOMINO study. Int J Clin Pract 69(1):33–40
Blasi F, Iori I, Bulfoni A, Corrao S, Costantino S, Legnani D (2008) Does implementation of pneumonia guidelines improve patient management? A three-phase study in Italian Internal Medicine departments. Eur Resp J 32:902–910
O’Brien MA, Rogers S, Jarntvedt G et al (2007) Educational outreach visits effects on professional practice and health care outcomes. Cochrane Database Syst Rev 17(4):Cd000409
Di Marco M, Miano N, Marchisello S et al (2023) Indirect Effects of the COVID-19 Pandemic on In-Hospital Outcomes among Internal Medicine Departments: A Double-Center Retrospective Study. J Clin Med 12(16):5304
Russo G, Di Bartolo P, Candido R et al (2023) The AMD ANNALS: A continuous initiative for the improvement of type 2 diabetes care. Diabetes Res Clin Pract 199:110672
Pandya N, Hames E, Sandhu S (2020) Challenges and Strategies for Managing Diabetes in the Elderly in Long-Term Care Settings. Diabetes Spectr 33(3):236–245
McCormack L, Sheridan S, Lewis M et al (2013) Communication and dissemination strategies to facilitate the use of health-related evidence. Evid Rep Technol Assess (Full Rep). ; (213):1–520
Ceriello A, Catrinoiu D, Chandramouli C et al (2021) Heart failure in type 2 diabetes: current perspectives on screening, diagnosis and management. Cardiovasc Diabetol 20(1):218
Prattichizzo F, de Candia P, Ceriello A (2021) Diabetes and kidney disease: emphasis on treatment with SGLT-2 inhibitors and GLP-1 receptor agonists. Metabolism 120:154799

Table 1. General characteristics of subjects recorded at admission to IMW in Phase 1 and in Phase 3 (EOV group and NO EOV group). #:Phase 1 vs Phase 3 no EOV; §: Phase 1 vs Phase 3 EOV; ¥: Phase 3 no EOV vs Phase 3 EOV. (Wilcoxon rank sum test)

VARIABLE	PHASE1 (n =1909)	PHASE 3		P value
VARIABLE	PHASE1 (n =1909)	EOV GROUP (n =1110)	NO-EOV GROUP (n= 552)	P value
Gender, n. (%)	1028 (53,8) M 881 (46,2) F	591 (53,2) M 519 (46,8) F	297 (53,8) M 255 (46,2)F	# 0.98 § 0.74 ¥ 0.82
Age, years mean ± SD	76,4 ±11,2	76,4 ± 10,5	76,0 ± 11,5	# 0.38 § 0.82 ¥ 0.51
Provenance of patients, n (%)
- home	1686 (88,3)	994 (89,5)	477 (86,4)	# 0.23 § 0.03 ¥ 0.002
- another dept	123 (6,4)	56 (5,0)	41 (7,4)
- nursing home	83 (4,3)	39 (3,5)	32 (5,8)
- long-term care facility	17 (0,9)	21 (1,9)	2 (0,4)
BMI, n (%)
- obese (>30)	405 (21,2)	248 (22,3)	112 (20,3)	# 0.58 § 0.30 ¥ 0.11
- overweight (25-30)	614 (32,2)	330 (29,7)	195 (35,3)
- normal weight (18,5-24,9)	807 (42,3)	493 (44,4)	223 (40,4)
- underweight (< 18,4)	83 (4,3)	39 (3,5)	22 (4,0)
Previous cardiovascular event (myocardial infarction, stroke, unstable angina, surgical revascularization), n (%)	779 (40,8)	418 (37,7)	245 (44,4)	# 0.13 § 0.09 ¥ 0.008
Concomitant diseases ≥3, n (%)	1064 (55,7)	555 (50,0)	313 (56,7)	# 0.69 § 0.002 ¥ 0.01
HbA1c (%) ± SD	7,2 ± 1,7	7,1 ± 1,3	7,2 ± 1,6	# 0.76 § 0.009 ¥ 0.12
Outcome, n (%)
Discharged	1713 (89,8)	998 (90,0)	467 (84,6)	#0.007 § 0.32 ¥ 0.001
Transferred to a nursing home or other clinical facilities	78 (4,1)	40 (3,6)	38 (6,9)
Transferred to another hospital	60 (3,1)	46 (4,1)	21 (3,8)
Transferred to another dept.	57 (3.0)	26 (2,3)	25 (4,5)
Actions specific for T2DM at the discharge, n (%)
referral to general practitioner	672 (35,2)	369 (33,2)	213 (38,6)	#0.001 § 0.002 ¥ 0.0002
no advice	395 (20,7)	179 (16,1)	98 (17,8)
referral to divisional ambulatory	(25,6)	345 (31,1)	137 (24,8)
referral to regional diabetes specialist	313 (16,4)	194 (17,5)	76 (13,8)
referral to other specialist consulting	41 (2,1)	23 (2,1)	28 (5,1)
Duration of hospitalization, median days (IQR)	9 (6-13)	9 (7-12)	10 (7-16)	#<0.0001 § 0.73 ¥ <0.0001

Table 2. Patients in therapy for T2DM in the two phases of the study. *(hypoglucidic diet, thiazolidinediones, acarbose). ∆% = (discharge - admission)/admission*100 (McNemar test).

	PHASE 1 (n=1909)			PHASE3 (n=1662)
	Admission	Discharge	P value	EOV GROUP (n=1110)		P value	NO EOV GROUP (n=552)		P value
	Admission	Discharge	P value	Admission	Discharge	P value	Admission	Discharge	P value
Patients in therapy for T2DM n.(%)	1559 (81,7)	1606 (84,21)	0,01	930 (83,8)	986 (88,82)	<0,0001	452 (81,9)	462 (83,7)	0,33
SGLT2i	34 (1,8)	53 (2,8)	0,008	69 (6,2)	234 (21,1)	<0,0001	15 (2,7)	58 (10,5)	<0,0001
GLP1ra	28 (1,5)	87 (4,6)	<0,0001	40 (3,6)	138 (12,4)	<0,0001	20 (3,6)	63 (11,4)	<0,0001
Metformin	1006 (52,7)	905 (47,4)	<0,0001	630 (56,8)	518 (46,7)	<0,0001	264 (47,8)	228 (41,3)	0,002
Metaglinides	191 (10,0)	150 (7,9)	0,0004	80 (7,2)	25 (2,2)	<0,0001	45 (8,1)	12 (2,2)	<0,0001
DPP4i	169 (8,8)	239 (12,5)	<0,0001	119 (10,7)	176 (15,9)	<0,0001	88 (15,9)	103 (18,7)	0,06
Sulfonylureas	247 (12,9)	140 (7,3)	<0,0001	118 (10,6)	51 (4,6)	<0,0001	48 (8,7)	19 (3,4)	<0,0001
Insulin	/	353 (18,5)	/	/	185 (16,7)	/	/	104 (18,8)	/
Combination	177 (9,3)	176 (9,2)	0,93	102 (9,2)	127 (11,4)	0,02	62 (11,2)	51 (9,2)	0,14
Other*	84 (4,4)	83 (4,3)	0,91	60 (5,4)	39 (3,5)	0,005	20 (3,6)	24 (4,3)	0,46

Table 3. Univariate and multivariable multilevel logistic analysis of predictors of non-adherence in Phase 3 population.

				Univariate			Multivariate
		N	Not Adherence *	OR	IC95%	Pvalue	OR	IC95%	Pvalue
Age, years	<75	636	319	Reference			Reference
Age, years	>=75	1026	714	2.1	1.6-2.6	<0.0001	2.0	1.6-2.6	<0.0001
Concomitant diseases, n.	<3	794	469	Reference
Concomitant diseases, n.	>=3	868	564	1.1	0.9-1.4	0.31
Obesity (BMI>30)	No	1302	834	Reference			Reference
Obesity (BMI>30)	Yes	360	199	0.7	0.5-0.9	0.003	0.7	0.5-0.9	0.04
Previous cardiovascular event	No	999	619	Reference
Previous cardiovascular event	Yes	663	414	1.1	0.9-1.4	0.32
Duration of hospitalization, days	<7	350	221	Reference
Duration of hospitalization, days	>=7	1312	812	1.0	0.7-1.3	0.88
FG at admission, mg/dL	<70	42	35	Reference		0.009	Reference		0.04
	70-125	549	371	0.3	0.1-0.8		0.4	0.2-0.9
	>125	1000	579	0.3	0.1-0.7		0.3	0.1-0.8

SupplementaryMaterial.docx

Management of Subjects with Type 2 Diabetes Hospitalized in Internal Medicine Units: A Cluster-Randomized, Multicenter Study Before & After an Educational Program

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods

3. Results

4. Discussion

5. Conclusions

Declarations

References

Tables

Supplementary Files

Status:

Version 1