2025+意大利指南：1型糖尿病的治疗

Abbreviations

AMD Associazione Medici Ospedalieri SID, Societa Italiana di Diabetologia PICOS Population, Intervention, Comparison, Out-

come, Study type

MNT Medical Nutrition Therapy

NPH Neutral Protamine Hagedorn

AMSTAR A MeaMeasurement Tool to Assess systematic Reviews

MH-OR Mantel-Haenzel Odds Ratio WMD Wehticed mean difference

GRADE of Recommendation, Assessment,Development, and Evaluation

EtD Evidence to Decision

RCTR Randomized Controlled Trial

OR Odds Ratio

CI Confidence Interval

RR Relative Risk

MD Mean Difference

QALY Quality Adjusted Life Years

ICER Incremental Cost-Effectiveness Ratio

BMI Body Mass Index

RT-CGM Real-Time Continuous Glucose Monitoring

isCGM Intermittently scanned Continuous Glucose

Monitoring

FGM 131 Flash Glucose Monitoring

MDI Multiple Daily Injections

CSII Continuous Subcutaneous Insulin Infusion

SAP Sensor Augmented Pump

TIR Time Range

MAGE Mean Amplitude of Glucose Excursion

Acta Diabetologica

Aims of the guideline

Type 1 diabetes mellitus is an autoimmune form of diabe-tes whose prevalence is rapidly increasing in both pediat-ric and adult settings[1]. In Italy, there are approximately 300,000 people with type 1 diabetes[2]. Type 1 diabetes is a disease of great influence on public health, with a con-siderable impact on disability and hospitalization rates and,consequently, with significant costs for healthcare systems.In Italy, the treatment of type 1 diabetes is entrusted to an integrated healthcare system that comprises a widespread network of specialist diabetes services (both pediatric and adult), supported by General Practitioners, Pediatricians of free choice and territorial healthcare facilities. This health-care system produces good quality care when compared to other European countries. However, important areas for improvement still remain: there is a fraction of people with type 1 diabetes who do not reach their therapeutic goals, and the management of insulin therapy, particularly technology-related therapy, still appears widely heterogeneous [3, 4].

A significant variability in treatment outcomes and costs lies in the use of technological instruments (i.e. insulin pumps, continuous glucose monitoring systems, and self-monitoring blood glucose devices) that have now become indispensable for the treatment of people with type 1 dia-betes. One of the reasons for this high heterogeneity is the rapid advancement of knowledges, which is difficult to syn-thesize in a rapid manner into clinical recommendations and to disseminate them among healthcare professionals.

The two main diabetes scientific societies in Italy (SID and AMD), with the participation of other healthcare pro-fessionals involved in the care of diabetes, have formulated the first joint guidelines on the treatment and management of type 1 diabetes in 2022. This clinical practice guideline aimed at providing a reference for pharmacological and non-pharmacological treatment of type 1 diabetes in individuals aged over 16 years. This guideline was primarily directed to physicians, nurses, dietitians and educators working in diabetes specialist clinics, as well as general practitioners,nurses and dietitians working in territorial services or pri-vate offices, and subjects affected by diabetes.

In this first update of the guideline, the panel has verified the need to modify, update, add or remove clinical ques-tions, and the opportunity of modifying the outcomes of interest and their relative relevance. In the case of changes in clinical questions and/or critical outcomes, the whole pro-cess of evidence reviews and development of recommenda-tions was performed anew. In all other cases, the guideline panel has reviewed and updated all systematic reviews by using the same search strings for each outcome of individ-ual question, verifying whether new evidence modified the risk/benefit ratio or the overall quality of evidences to the

extent of modifying the formulation of a recommendation,its strength or its quality of evidence.

In this first update of the guideline, the following specific areas were assessed: therapeutic goals, nutritional therapy,physical exercise, educational programs, psychological support, pharmacological treatment, glucose monitoring,use of technology, and pancreatic islet transplantation. All the interventions considered are usually reimbursed in Italy,with some regional differences for continuous insulin infu-sion devices, glucose monitoring devices and nutritional therapy. The recommendations presented in this update of the guideline have been formulated based on available evi-dence, independent of current reimbursement policies, and have been designed as indications for healthcare profession-als in charge of diabetes treatment, primarily based on clini-cal needs of people with type 1 diabetes and considering the existing organization of healthcare. These recommenda-tions apply to outpatients with type 1 diabetes(aged over 16 years), either in primary care or at specialist referral.

The implementation of this updated Guideline will be pursued through its dissemination, performed by:(1) Scien-tific Societies, using their websites and official journals and organizing specific activities of continuous medical educa-tion; and(2) Regional healthcare systems.

Methods for guideline development

The present Guideline was developed following the meth-ods described in the Manual of the National Guideline Sys-tem(http://www.snlg-iss.it).

Clinical questions

Each recommendation answers a clinical question, formu-lated by the panel using the PICOS framework.

Selection of outcomes

For each question, the panel identified potentially relevant outcomes, which were then rated for their impact on thera-peutic choices using a 9-point scale:

$0-3$ points: outcomes of limited relevance

4-6 points: important, but not critical outcomes

7-9 points: critical outcomes.

Only outcomes classified as "critical" were considered in the systematic review of evidences and in the formulation of recommendations.

Acta Diabetologica

Evidence review and assessment of quality of evidence

A systematic review for critical outcomes for each question was performed on the following databases:

● Cochrane Database of Systematic Reviews(Wiley).

●●Cochrane Central Register of Controlled Trials (Wiley).

● MEDLINE(OVID).

● Embase (OVID).

● Clinicaltrials.gov.

Specific search strategies were used for each database, as specified in each chapter of the Appendix. Only items in English were considered. References of retrieved items were searched for further studies meeting inclusion criteria.

The systematic review was performed through the fol-lowing steps:

● Selection of potentially eligible studies obtained with the initial search, on the basis of title and abstract, for retrieval as full text;

● Identification among retrieved full-text items of relevant studies, on the basis of a priori inclusion and exclusion criteria;

● Critical assessment of the risk of bias using validated in-struments (i.e., AMSTAR 2 for systematic reviews and the Cochrane collaboration tool for randomized trials);

● Extraction of the main characteristics of selected stud-ies(population enrolled, considered outcomes, results), summarized in tables;

● Quantitative synthesis for each outcome, calculating MH-OR for categorial outcomes and WMD for continu-ous variables, both with 95% confidence intervals. The main analysis was always performed with random ef-fects models, whereas fixed effects models, when used,were considered only for sensitivity analyses;

● Assessment of heterogeneity $\left(I^{2}\right)$ and of publication bias(Funnel plot);

● The overall quality and strength of available evidence for outcomes selected by the panel were rated using the GRADE ${ }^{5}$ criteria;

● Synthesis of results, using the GRADEPro Guideline Development tool (https://gradepro.org), with the fram ews $EtD^{6}$ , which summarize results of systematic re-views for problem priority, desired and undesired effects of treatments, strength of available evidence, values and preferences of stakeholders, economic resources need-ed, equity, acceptability and feasibility of interventions

(for the full GRADE assessment tables please refer to supplementary material).

Statistical analyses were performed with RevMan 5.0 (https://training.cochrane.org/online-learning/core-software-cooch rane-reviews/revman/revman-5-download).

For pharmacoeconomic studies, relevant records were selected on the basis of title and abstract for full text retrieval. Due to the geographical and methodological het-erogeneity of retrieved studies, no formal meta-analysis was performed; methods and results were summarized in tables,including type of analysis, context, year(s) to which costs were referred, efficacy, cost-efficacy and cost-utility, main conclusions.

Development of recommendations

The guideline panel has examined and discussed, for each clinical question, EtD frameworks, tables of evidence and summaries of results(i.e. forest plots of meta-analyses).Recommendations have been formulated on the basis of results of published studies and level of evidence. Disagree-ments were resolved through collective discussion.

External review

The guideline panel has identified three external reviewers,chosen among Italian healthcare professionals with a spe-cific experience of clinical research in type 1 diabetes, with known methodological skills, who had published at least 150 peer-reviewed original articles on International medi-cal journals and who had a H-index of at least 30. Members of the guideline panel and evidence review team, and cur-rent members of the national board of SID or AMD, were excluded.

External reviewers received a draft version of the guide-line and provided their observations to the panel. The panel collectively discussed the points raised by the external reviewers, elaborating the amendments to the guideline and the response to reviewers.

Guideline update

Systematic reviews will be updated, using the same search strings, once every year, starting from the date of final approval of the guideline. The evidence review team and the guideline panel will verify whether new evidence will mod-ify the risk/benefit ratio or the overall quality of evidences

Acta Diabetologica

to the extent of modifying the formulation of a recommen-dation, its strength or its quality of evidence.

Once every year, the Guideline panel will verify the need to modify, update, add or remove clinical questions,and the opportunity of modifying the outcomes of interest and their relative relevance. In the case of changes in clini-cal questions and/or critical outcomes, the whole process of evidence review and development of recommendations will be performed anew.

Interpretation of recommendations

Quality of evidence

HIGH: Highly reliable results. It is very unlikely that further studies modify the confidence in estimated effects.

MODERATE: Moderately reliable results. It is possible that further studies modify the confidence in estimated effects.

LOW: Results are still uncertain. Further research is needed for a reliable assessment of positive and negative effects of the intervention.

VERY LOW: Available data are not reliable, and esti-mates of effects should be considered with caution.

Strength of recommendations

Strong recommendation

● for clinicians: the majority of patients must receive the recommended interventions;

● for patients: almost all properly informed patients fol-low the recommendation and only a small proportion chose different options;

● for policy makers: the recommendation can be used for planning the use of available resources.

Weak recommendation

● for clinicians: the final choice should include a careful consideration of the patient's values and preferences;

● for patients: the majority of properly informed patients follow the recommendation, but a minority chose differ-ent options;

● for policy makers: a discussion involving stakeholder should be developed.

Summary of recommendations

1. Pharmacological treatment

1.1. We recommend the use of basal-bolus insulin regimens, instead of premixed insulin,for patients with type 1 diabetes.

Strength of the recommendation: weak. Quality of evidence: low.

1.2. We recommend the use of a long-acting insulin analogue, instead of NPH human insulin, in subjects with type 1 diabetes mellitus on basal-bolus insulin regimens.

Strength of the recommendation: strong. Quality of evidence: moderate.

1.3. We recommend the use of a short-acting insulin analogue as prandial insulin in subjects with type 1 diabetes mellitus on intensive insulin therapy.

Strength of the recommendation: strong. Quality of evidence: moderate.

1.4. We suggest the use of an insulin needle no longer than 4 mm in subjects with type 1 diabetes mellitus on multiple daily injection insulin therapy.

Acta Diabetologica

Strength of the recommendation: weak. Quality of evidence: moderate.

1.5. We recommend to offer the patient the option of insulin pump rather than multiple daily injection insulin therapy in subjects with inadequately controlled type 1 diabetes mellitus because of the advantages on glycated hemoglobin levels, severe hypoglycemia,quality of life and treatment satisfaction.

We suggest to offer the patient the option of insulin pump rather than multiple daily injection insulin therapy in subjects with controlled type 1 diabetes mellitus because of the advantages on glycemic variability, quality of life and treatment satisfaction.

Strength of the recommendation: strong. Quality of evidence: moderate.

1.6. We recommend the use of closed loop systems consisting of an insulin pump and a glycemic sensor with automatism instead of systems without automatism in subjects with type 1 diabetes mellitus.

Strength of the recommendation: strong. Quality of evidence: high.

2. Glucose monitoring

2.1. We recommend the use of continuous glucose monitoring system instead of self-monitoring of blood glucose in subjects with type 1 diabetes mellitus.

Strength of the recommendation: strong. Quality of evidence: moderate.

2.2. We recommend the use of continuous glucose monitoring systems with predictive alerts rather than systems without predictive alerts in subjects with type 1 diabetes mellitus with a history of severe hypoglycemia or hypoglycemia unawareness or with frequent episodes of hypoglycemia.

Strength of the recommendation: strong. Quality of evidence: moderate.

3.Physical exercise

3.1. There is no evidence to prefer a specific type of physical activity (i.e. exclusively aerobic or against-resistance) in subjects with type 1 diabetes.

Strength of the recommendation: weak. Quality of evidence: moderate.

4.Nutritional therapy

4.1. We suggest to use carbohydrate counting technique to assess the exact amount of insulin to be administered at mealtimes in subjects with type 1 diabetes mellitus.

Strength of the recommendation: weak. Quality of evidence: low.

Acta Diabetologica

4.2. We suggest to use a diet that is not exclusively based on low-glycemic index nutrients in subjects with type 1 diabetes mellitus.

Strength of the recommendation: weak. Quality of evidence: very low.

5. Treatment targets

5.1. We recommend an intensive glycemic control, with a target of HbA1c between 48 and 53 mmol/mol(6.5%-7.0%) in subjects with type 1 diabetes mellitus, in order to reduce the risk of microangiopathic complications, cardiovascular disease and hypoglycemia.

Strength of the recommendation: strong. Quality of evidence: high.

5.2. A target Time ln Range (glucose levels between 70 and 180 mg/dl) of 70\% or more is recommended for subjects with type 1 diabetes mellitus for the benefits on glycated hemoglobin levels, the degree of glycemic variability, the onset and/or progression of microangiopathic complications.

6.1. We suggest starting evaluation for pancreatic islet transplantation in subjects with unstable type 1 diabetes mellitus, despite the best optimized possible insulin therapy.

Strength of the recommendation: weak. Quality of evidence: moderate.

7.Educational therapy

7.1. We suggest the use of structured educational programs rather than providing information in an unstructured manner in subjects with uncontrolled type 1 diabetes mellitus.

Strength of the recommendation: weak. Quality of evidence: moderate.

7.2. We suggest to implement a structured transition process from pediatric to adult diabetology in subjects with type 1 diabetes mellitus followed in a pediatric care setting.

Acta Diabetologica

1.1 Insulin regimens

Question: Should basal-bolus insulin regimens be preferred to premixed insulin in patients with type 1 diabetes?

|Population|People with type 1 diabetes| |---|---| |Intervention|Basal-bolus insulin regimen| |Comparison|Premixed insulin| |Outcome|HbA1c,Hypoglycemia, Quality of life,Treatment satisfaction| |Setting|Outpatient|

Recommendation:

We recommend the use of basal-bolus insulin regimens,instead of premixed insulin, for patients with type 1 diabetes.

Strength of the recommendation: weak. Quality of evi-dence: low.

Justification. Several randomized trials show that basal-bolus insulin regimens are associated with lower risk of hypoglycemia and better treatment satisfaction than premixed insulin. Even if the reduction in HbA1c levels was not significant, the use of basal-bolus schemes was cost-effective.

Although the available literature on this topic is dated,the evidence supports this recommendation due to consis-tence of findings and biological plausibility.

Subgroup considerations. There are no available data on safety and efficacy in subgroups of patients with well-controlled, poorly controlled diabetes or a history of severe hypoglycemia, probably due to lacking trials including patients treated with premixed insulin.

Where heterogeneity between studies in the meta-anal-ysis performed for critical outcomes was detected, it was decided that this would be explored using the subgroups predefined in the protocol.

HbAlc: subgroup analysis was performed according to the type of insulin used in each study. This analysis showed that the use of different insulin types did not explain the heterogeneity between the studies.

Insulin dose or regimen: all studies used different insu-lins in the basal-bolus scheme. The different insulins and regimens used could be a possible explanation for the het-erogeneity among the studies.

Age of participants: the average age of participants in the trials included in the meta-analysis was very similar, this would not explain any heterogeneity between the studies.

Hypoglycemia at baseline: Even when studies exclud-ing subjects with a history of severe hypoglycemia were removed from the meta-analysis, the heterogeneity remained statistically significant.

Implementation. Specific programs for continuous medical education are helpful to increase the awareness of healthcare professionals of the benefits of basal-bolus

<EndOfPara>insulin regimens and, conversely, the disadvantages associ-ated with the use of premixed insulin.

For subjects with suboptimal blood glucose control and wide glycemic fluctuations, it is necessary to check injec-tion technique, injection site, self-monitoring capacity,knowledge and self-management abilities, as well as life-style, psychological and psychosocial difficulties, and pos-sible organic causes, such as gastroparesis.

Guidelines and clear protocols should be provided to all adults with type 1 diabetes to help them regulate insulin doses appropriately during periods of disease.

Assessment and monitoring. Adherence to this guideline can be assessed by estimating the proportion of subjects treated with premixed insulin out of the total number of sub-jects on insulin therapy in existing databases ${}^{3,4}$ .

1.2 Long-acting insulin analogues

Question: Is it preferable to use long-acting insulin ana-logues or human NPH insulin as basal insulin for subjects with type 1 diabetes mellitus who use basal-bolus insulin schemes?

|Population|People with type 1 diabetes| |---|---| |Intervention|Long-acting insulin analogues| |Comparison|Human NPH insulin| |Outcome|Severe Hypoglycemia| |Setting|Outpatient|

Recommendation:

In subjects by 1 diabetes mellitus who use basal-bolus insulin schemes it is recommended to use as basal insulin a long-acting insulin analogue compared to human NPH insulin.

Strength of the recommendation: strong. Quality of evi-dence: moderate.

Justification. Several randomized clinical trials have highlighted advantages on the risk of hypoglycemia when participating subjects used basal-bolus insulin schemes,including long-acting insulin analogues versus insulin schemes containing NPH insulin. The use of long-acting insulin analogues has been found to be cost-effective.Although the available literature on this topic is dated, the evidence supports this recommendation due to consistence of findings and biological plausibility.

Subgroup considerations. Where heterogeneity in meta-analyses performed for critical outcomes (in such case, only hypoglycemia) was found, it was agreed that this would be explored using the predefined subgroups in the protocol.

Baseline HbAlc: could not be explored because HbAlc levels were similar in most studies(between 7.5% and 8.5%) and in none of studies was lower than 7.0%.

Insulin dose or regimen: Regarding the insulin dose, there was no heterogeneity as all studies used a strict insulin titra-tion protocol. Regarding the insulin regimen, some studies have used insulin schemes including the same insulins but

Acta Diabetologica

with different numbers of administrations per day (detemir once or twice a day; NPH mostly twice a day). In the com-parison of detemir versus NPH, the insulin regimen hetero-geneity between studies increased. In the glargine versus NPH comparison, the insulin regimen explained the hetero-geneity between studies.

Age of participants: the average age of participants in the trials included in the meta-analysis was very similar(35-45 years), which would not explain any heterogeneity between the studies.

Baseline hypoglycemia: Not reported in most studies.

Baseline weight: it was not possible to explore this data as there was evidence of heterogeneity in the findings. In fact, in some studies, only the weight expressed in kg was present and in other studies, the BMI was present.

Implementation. Specific programs for continuous medical education should be planned to increase the aware-ness of healthcare professionals of the benefits of long-act-ing insulin analogues.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of subjects treated with basal insulin analogues out of the total number of subjects on basal insulin therapy in existing databases3,4.

1.3 Prandial insulin

Question: Is it preferable to use short-acting insulin ana-logues or human regular insulin for subjects with type 1 dia-betes mellitus on intensive insulin therapy?

|Population|People with type 1 diabetes| |---|---| |Intervention|Short-acting insulin analogues| |Comparison|Human regular insulin| |Outcome|Hypoglycemia, HbAlc| |Setting|Outpatient|

Recommendation:

We recommend the use of a short-acting insulin analogue as prandial insulin in subjects with type 1 diabetes mellitus on intensive insulin therapy.

Strength of the recommendation: strong. Quality of evi-dence:moderate.

Justification. Several randomized clinical trials have highlighted advantages on the risk of hypoglycemia and HbA1c when participating subjects used basal-bolus insu-lin schemes, including short-acting insulin analogues ver-sus insulin schemes containing human regular insulin. The use of short-acting insulin analogues has been found to be cost-effective. Although the available literature on this topic is dated, the evidence supports this recommendation due to consistence of findings and biological plausibility.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes (in this case, HbAlc and hypoglycemia), there-fore, the subgroups predefined in the protocol were not explored.

Implementation. Specific programs for continuous medical education are helpful to increase the awareness of healthcare professionals of the benefits of short-acting insu-lin analogues use and, conversely, the disadvantages associ-ated with the use of regular human insulin.

For subjects with suboptimal blood glucose control and characterized by wide glycemic fluctuations, it is necessary to check injection technique, injection site, self-monitoring capacity, knowledge and self-management abilities, life-style, as well as psychological and psychosocial difficulties,and possible organic causes, such as gastroparesis.

Guidelines and clear protocols should be provided to all adults with type 1 diabetes to help them manage insulin doses appropriately during periods of disease.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of sub-jects treated with rapid insulin analogues out of the total number of subjects on prandial insulin therapy in existing databases,3,4

1.4 Insulin needles

Question: Is it preferable to use a short or long insulin nee-dle for subjects with type 1 diabetes mellitus on basal-bolus insulin schemes?

|Population|People with type 1 diabetes| |---|---| |Intervention|Short insulin needle| |Comparison|Long insulin needle| |Outcome|Hypoglycemia, lipodystrophies,pain| |Setting|Outpatient|

Recommendation:

We suggest the use of an insulin needle no longer than 4 mm in subjects with type 1 diabetes mellitus on multi-injection insulin therapy.

Strength of the recommendation: weak. Quality of evi-dence:moderate.

Justification. Several randomized clinical trials have highlighted advantages on pain symptoms during insulin administration when participants used short insulin needles compared to longer needles. Nevertheless, there is no evi-dence of an advantage in terms of a significant reduction in the risk of hypoglycemic episodes. This results in a weak strength of recommendation despite a moderate quality of evidence. The use of short insulin needles was found to be cost-effective.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes.

Implementation. Specific programs for continuous med-ical education are useful to increase the awareness of health-care professionals of the benefits of short insulin needles.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of sub-jects using short insulin needles out of the total number of

Acta Diabetologica

insulin-treated subjects in existing dedicated pharmaceuti-cal databases.

Research priorities. Clinical studies comparing needles with different characteristics (e.g. needle geometry) are desirable.

1.5 Insulin administration methods

Question: Is it preferable to use an insulin pump rather than multiple daily injection insulin therapy for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Insulin pump| |Comparison|Multiple daily injection insulin therapy| |Outcome|Hypoglycemia, HbA1c, glycemic variability, quality of life, treatment satisfaction| |Setting|Outpatient|

Recommendation:

We recommend offering the patient the option of an insu-lin pump therapy rather than multiple daily injection insulin therapy in subjects with poorly controlled type 1 diabetes mellitus because of the advantages on glycated hemoglobin levels, severe hypoglycemia, quality of life and treatment satisfaction.

We suggest offering the patient the option of an insulin pump therapy rather than multiple daily injection insulin therapy in subjects with controlled type 1 diabetes mellitus because of the advantages on glycemic variability, quality of life and treatment satisfaction.

Strength of the recommendation: strong. Quality of evi-dence: moderate.

Justification. Several randomized clinical trials have highlighted advantages of insulin pump therapy for glycated hemoglobin levels, glycemic variability, severe hypoglyce-mia, quality of life and treatment satisfaction when partici-pating subjects used insulin pumps compared to multiple daily insulin injections. The use of insulin pumps has been found to be cost-effective.

Subgroup considerations. Separate analyses were con-ducted for three distinct subgroups of patients(subjects with poorly controlled diabetes, subjects with controlled diabe-tes, and subjects with a history of severe hypoglycemia).For each of the subgroups, the effects on critical outcomes were explored.

Specific categories of patients(psychiatric patients with cognitive impairment or sensory deficits) are routinely excluded from clinical trials. For such patients, careful eval-uation is necessary before starting such therapy.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of insulin pump use and, conversely, the disadvantages associated with multiple daily injection insulin therapy.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of subjects treated with insulin pumps out of the total number of sub-jects on insulin therapy in the three distinct subgroups of patients(subjects with poorly controlled diabetes, subjects with controlled diabetes, subjects with a history of severe hypoglycemia) in existing databases3.4.

Research priorities. Studies on barriers to the choice of pump use by both the person with diabetes and the health care professional could be useful.

1.6 Automated insulin delivery systems

Question: Is it preferable to use closed-loop systems con-sisting of an insulin pump and glucose sensor with automa-tism or systems without automatism for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Closed-loop systems consisting of insulin pump and glucose sensor with automatism| |Comparison|Systems consisting of insulin pump and glucose sensor without automatism| |Outcome|Hypoglycemia, HbA1c, glycemic vari- ability, Time In Range (TIR), quality of life, treatment satisfaction| |Setting|Outpatient|

Recommendation:

We recommend the use of closed-loop systems consist-ing of insulin pump and glucose sensor with automatism instead of systems without automatism in subjects with type 1 diabetes mellitus.

Strength of the recommendation: strong. Quality of evi-dence:high.

Justification. Several randomized clinical trials have highlighted advantages on HbAlc, Time in Range and gly-cemic variability when participating subjects used closed-loop systems consisting of pump and glucose sensors with automatism compared to systems without automatism. The use of systems consisting of an insulin pump and glucose sensor with automatism has been found to be cost-effective.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes, therefore, the subgroups predefined in the proto-col were not explored.

Implementation. Specific programs for continuous medical education are helpful to increase the awareness of healthcare professionals of the benefits of using systems consisting of an insulin pump and glucose sensor with automatism.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of sub-jects treated with insulin pumps and glucose sensors with automatism out of the total number of subjects treated with insulin pumps in existing databases ${ }^{3,4}$ .

Acta Diabetologica

Research priorities. Studies on barriers to the choice of closed-loop systems use by both the person with diabetes and the health care professional could be useful.

2. Glucose monitoring

2.1 Continuous glucose monitoring

Question: Is it preferable to use continuous glucose moni-toring systems or self-monitoring of blood glucose(SMBG)for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Continuous Glucose Monitoring systems| |Comparison|Self-Monitoring of Blood Glucose(SMBG)| |Outcome|Hypoglycemia, HbA1c, Glycemic vari- ability, Time in Range, Quality of Life, Treatment satisfaction| |Setting|Outpatient|

Recommendation:

We recommend the use of a continuous glucose monitor-ing system instead of self-monitoring of blood glucose in subjects with type 1 diabetes mellitus.

Strength of the recommendation: strong. Quality of evi-dence: moderate.

Justification. Several randomized clinical trials have highlighted advantages on hypoglycemia risk, glycated hemoglobin levels, Time In Range and glycemic variabil-ity when participating subjects used continuous glucose monitoring compared to self-monitoring of blood glu-cose. Continuous glucose monitoring has been found to be cost-effective.

Subgroup considerations. No predefined subgroups were considered.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of continuous glu-cose monitoring use and, conversely, the disadvantages associated with self-monitoring of blood glucose.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of sub-jects treated with continuous glucose monitoring out of the total number of subjects with type 1 diabetes in existing databases ${ }^{3,4}$ .

Research priorities. None.

2.2 Predictive alerts of continuous glucose monitoring Question: Is it preferable to use continuous glucose moni-toring systems with predictive alerts or systems without predictive alerts for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Continuous glucose monitoring| |Comparison|systems with predictive alerts| |Comparison|Continuous glucose monitoring| |Outcome|using hypothesis| |Outcome|hypoglycemia, HbA1c, Glycemic variability, Time in Range, Qual-ity of Life, Treatment satisfaction|

|Population|People with type 1 diabetes| |---|---| |Setting|Outpatient|

Recommendation:

It is recommended to use continuous glucose monitoring systems with predictive alerts rather than systems without predictive alerts in subjects with type 1 diabetes mellitus with a history of severe hypoglycemia or hypoglycemia/unawareness or with frequent episodes of hypoglycemia.

Strength of the recommendation: strong. Quality of evi-dence: moderate.

Justification. Available evidence has shown advantages on severe hypoglycemia risk, Time In Range and time below range when participating subjects used continuous glucose monitoring with predictive alerts compared to systems with-out predictive alerts.

Subgroup considerations. No predefined subgroups were considered.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of the use of con-tinuous glucose monitoring with predictive alerts.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of subjects treated with continuous glucose monitoring with predictive alerts out of the total number of subjects with type 1 diabe-tes in existing databases ${ }^{3,4}$ .

Research priorities. Specific cost-effectiveness analyses should be performed.

3. Physical exercise

3.1 Physical exercise

Question: Is it preferable to exercise exclusively aerobi-cally or against-resistance for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Exclusively aerobic physical activity| |Comparison|Against-resistance physical activity| |Outcome|Hypoglycemia, HbA1c, Body weight| |Setting|Outpatient|

Recommendation:

There is no evidence to prefer a specific type of physical activity (exclusively aerobic or against-resistance) in sub-jects with type 1 diabetes.

Strength of the recommendation: weak. Quality of evi-dence: moderate.

Justification. Several randomized clinical trials have highlighted no advantages or disadvantages in terms of hypoglycemia risk, glycated hemoglobin and body weight when participating subjects carried out exclusively aerobic physical activity rather than against-resistance. No cost-effectiveness evaluations are available. This results in a weak strength of recommendation despite a moderate qual-ity of evidence.

Acta Diabetologica

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes, therefore no subgroups were explored.

Implementation. Specific programs for continu-ous medical education are helpful to increase the aware-ness of healthcare professionals of the neutrality of performing exclusively aerobic physical activity rather than against-resistance.

Assessment and monitoring. Monitoring and evalua-tion of guideline adherence concerning recommendations on non-pharmacological interventions is problematic. The inclusion in the medical record of information regarding recommendations on type of physical activity and dura-tion to be used as indicators of quality of care provided by general practitioners and diabologists could be useful to evaluate the implementation of this recommendation.

Research priorities. Specific cost-effectiveness analyses should be performed.

4. Nutritional therapy

4.1 Carbohydrate counting

Question: Is using the carbohydrate counting technique ben-eficial for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Carbohydrate counting| |Comparison|Pardial insulin dose without carbohydrate counting| |Outcome|Hypoglycemia,Hyperglycemia,HbA1c| |Setting|Outpatient|

Recommendation:

We suggest to use the carbohydrate counting technique to assess the exact amount of insulin to be administered at mealtimes in subjects with type 1 diabetes mellitus.

Strength of the recommendation: weak. Quality of evi-dence: low.

Justification. Several randomized clinical trials have highlighted advantages on glycated hemoglobin levels when participating subjects used carbohydrate counting compared to the dosage of prandial insulin without carbo-hydrate counting.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes, therefore, no subgroups were explored.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of the use of carbo-hydrate counting. The presence within the diabetes team of professionals with expertise in nutrition eases the achieve-ment of the goals provided by this recommendation.

Assessment and monitoring. Monitoring and evalua-tion of guideline adherence with regard to recommendations on non-pharmacological interventions is problematic. The inclusion in the medical record of information regarding recommendations on the type of diet to be used as indicators

<EndOfPara>of quality of care provided by general practitioners and dia-betologists could be useful to evaluate the implementation of this recommendation.

Research priorities. Clinical studies with large popula-tion and multiple outcomes are desirable.

4.2 Type of diet

Question: Is using a low or high glycemic index diet prefer-able for subjects with type 1 diabetes mellitus?

|Implementation|People with type 1 diabetes| |---|---| |Intervention|Low glycemic index diet| |Comparison|Other types of diet| |Outcome|Hypoglycemia, Hyper-| ||glycemia, HbAlc, Body| |Setting|Outpatient|

Recommendation:

We suggest to use a diet that is not exclusively based on low-glycemic index nutrients in subjects with type 1 diabe-tes mellitus.

Strength of the recommendation: weak. Quality of evi-dence: very low.

Justification. Several randomized clinical trials have shown a neutral effect on glycated hemoglobin levels, risk of hypoglycemia and hyperglycemia, body weight, lipid profile, and patient preference when participating subjects consumed low glycemic index foods compared to high gly-cemic index foods. Therefore, there is insufficient evidence to recommend an exclusively low-glycemic index diet over other healthy dietary patterns.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes, therefore no subgroups were explored.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the neutrality of the intake of low glycemic index foods compared with the intake of foods with a high glycemic index.

Assessment and monitoring. Monitoring and evalua-tion of guideline adherence about recommendations on non-pharmacological interventions is problematic. The inclusion in the medical record of information regarding recommen-dations on the type of diet to be used as indicators of quality of care provided by general practitioners and diabetolo-gists could be useful to evaluate the implementation of this recommendation.

Research priorities. Clinical studies with large popula-tion and multiple outcomes are desirable.

5. Treatment targets

5.1 Glycated hemoglobin

Question: Is intensive glycemic control more effec-tive in reducing the risk of microvascular complications,

Acta Diabetologica

cardiovascular disease, and hypoglycemic events compared with standard control for subjects with type 1 diabetes mellitus?

|Population|People with type 1 diabetes| |---|---| |Intervention|Intensive glycemic control| |Comparison|Standard glycemic control| |Outcome|Hypoglycemia, Micro- vascular complications, Cardiovascular disease| |Setting|Outpatient|

Recommendation:

An intensive glycemic control, with a HbA1c target between 48 and 53 mmol/mol (HbA1c 6.5-7.0%), is recom-mended for subjects with type 1 diabetes mellitus, in order to reduce the risk of microangiopathic complications, car-diovascular disease and hypoglycemia.

Strength of the recommendation: strong. Quality of evi-dence: high.

Justification. Several randomized clinical trials have highlighted advantages on microvascular complications and cardiovascular disease when participating subjects fol-lowed an intensive glycemic control. Intensive glycemic control was found to be cost-effective. HbA1c targets below 48 mmol/mol or 6.5%, while reducing the risk of microvas-cular complications have, on the other hand, an unfavorable risk-benefit ratio because of the increased risk of severe hypoglycemia.

Subgroup considerations. Efficacy and safety data are not available in subjects with more than 70 years of age,limited life expectancy, history of frequent severe hypo-glycemia, significant comorbidities. These considerations would lead to the recommendation of higher HbA1c values in these categories of subjects.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of intensive glyce-mic control.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of subjects reaching the HbA1c target in existing databases ${ }^{3,4}$ .

5.2 Time in range

Question: For subjects with type 1 diabetes mellitus, is a Time In Range of 70\% or greater more effective than a Time In Range less than 70\% in improving HbA1c levels, the degree of glycemic variability and in preventing the onset and/or progression of microangiopathic complications?

|Population|People with type 1 diabetes| |---|---| |Intervention|Time In Range of 70% or greater| |Comparison|Time In Range less than 70%| |Outcome|Hypoglycemia, HbA1c,Gly- cemic variability, Microangio- pathic complications,| |2000|cardiovascular disease|

|Population|People with type 1 diabetes| |---|---| |Setting|Outpatient|

Recommendation:

A target Time In Range (glucose levels between 70 and 180 mg/dl) of 70% or more is recommended for subjects with type 1 diabetes mellitus for the benefits on glycated hemoglobin levels, the degree of glycemic variability, the onset and/or progression of microangiopathic complications.

Strength of the recommendation: strong. Quality of evi-dence: moderate.

Justification. Only one randomized, but large, clinical trial showed benefits on microvascular complications(par-ticularly diabetic retinopathy) when participating subjects achieved and maintained a Time In Range (glucose levels between 70 and 180 mg/dl) of 70% or higher.

Subgroup considerations. Efficacy and safety data are not available in subjects over 70 years of age, subjects with limited life expectancy, in subjects with a history of fre-quent severe hypoglycemia, and in subjects with significant comorbidities.

These considerations would lead to the recommendation of lower TIR values in these categories of subjects.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of a TIR of 70% or higher.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of subjects reaching and maintaining the target Time In range in exist-ing databases ${ }^{3,4}$ .

6. Pancreatic islet transplantation

6.1 Pancreatic islet transplantation

Question: In subjects with unstable type 1 diabetes mellitus,what are the indications for pancreatic islet transplantation?

|Population|People with type 1 diabetes| |---|---| |Intervention|Pancreatic islet transplantation| |Comparison|Intensive insulin therapy| |Outcome|Hypoglycemia, HbA1c,| |Setting|Microvascular complications,| |Setting|Cardiovascular disease, Qual-| |Setting|ity of life| |Outpatient||

Recommendation:

We suggest starting evaluation for pancreatic islet trans-plantation in subjects with unstable type 1 diabetes mellitus,despite the best optimized insulin therapy possible.

Strength of the recommendation: weak. Quality of evi-dence: moderate.

Justification. Unstable diabetes consists of severe glyce-mic lability associated with at least two episodes of severe hypoglycemia per year, severe impairment of quality of life related to hypoglycemia or hypoglycemia unawareness in

Acta Diabetologica

subjects with a BMI of less than 30 kg/m2, a daily insulin requirement of less than 0.85 units per kg, and unmeasur-able levels of c-peptide.

Several randomized clinical trials have shown benefits on glycated hemoglobin levels, the risk of hypoglycemia and development and/or progression of microangiopathic complications when subjects participants had undergone pancreatic islet transplantation. While clinically beneficial for a very select, highly unstable group, its high cost and unfavorable cost-effectiveness make this a weak recommen-dation reserved for specific scenarios where other options have failed and potential benefits are deemed to outweigh these factors.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes, therefore, the subgroups predefined in the proto-col were not explored.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of pancreatic islet transplantation.

Assessment and monitoring. Adherence to this guide-line can be assessed by estimating the proportion of subjects undergoing pancreatic islet transplantation or sent for evalu-ation of.

transplantation.

Research priorities. Research into the best technique for performing the transplant could reduce the undesirable effects associated with the immunosuppressive therapy cur-rently used to prevent rejection.

7. Educational therapy and transition

7.1 Structured educational therapy

Question: Is the use of structured therapeutic education programs more effective than providing information in an unstructured way in patients with type 1 diabetes?

|Population|People with type 1 diabetes| |---|---| |Intervention|Structured therapeutic education programs| |Comparison|Unstructured standard education| |Outcome|Hypoglycemia, HbA1c, Quality of life| |Setting|Outpatient|

Recommendation:

We suggest the use of structured educational programs rather than providing information in an unstructured manner in subjects with uncontrolled type 1 diabetes mellitus.

Strength of the recommendation: weak. Quality of evi-dence: moderate.

Justification. Several randomized clinical trials have shown an advantage on glycated hemoglobin levels but not on the risk of hypoglycemia and quality of life when structured therapeutic education programs were initiated for participating subjects. This results in a weak strength

<EndOfPara>of recommendation despite a moderate quality of evidence.The intervention has been shown to be cost-effective.

Subgroup considerations. Subgroups consisting of patients with controlled diabetes and those with uncon-trolled diabetes were explored.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits of structured educa-tional programs.

Assessment and monitoring. Monitoring and evalua-tion of guideline adherence with regard to recommendations on non-pharmacological interventions is problematic. The inclusion in the medical record of information regarding the type of structured educational programs as indicators of quality of care provided by general practitioners and diabe-tologists could be useful in order to evaluate the implemen-tation of this recommendation.

7.2 Pediatric-Adult Transition

Question: For adolescents with type 1 diabetes mellitus, is it useful to perform a structured transition process from pedi-atric to adult diabetology?

|Population|People with type 1 diabetes| |---|---| |Intervention|Transition from pediatric to adult diabetology| |Comparison|No transition| |Outcome|Hypoglycemia, HbAlc, Quality of life| |Setting|Outpatient|

Recommendation:

We suggest to realize a structured transition process from pediatric to adult diabetology in subjects with type 1 diabe-tes mellitus followed in a pediatric care setting.

Strength of the recommendation: weak. Quality of evi-dence: moderate.

Justification. Several randomized clinical trials have shown an advantage on the risk of hypoglycemia but no advantage on glycated hemoglobin levels and quality of life when subjects were involved in a structured transition from pediatric to adult diabetes. No evaluations of cost-effective-ness exist. This results in a weak strength of recommenda-tion despite a moderate quality of evidence.

Subgroup considerations. No heterogeneity was found among studies in the meta-analyses performed for critical outcomes, therefore no subgroups were explored.

Implementation. Specific programs for continuous medical education are useful to increase the awareness of healthcare professionals of the benefits associated with the structured transition process from pediatric to adult diabetology.

Assessment and monitoring. Monitoring and evalua-tion of guideline adherence with regard to this recommen-dation involves analyzing the number of subjects with type 1 diabetes treated by adult diabetes services in relation to the total number of subjects treated.

Acta Diabetologica

Research priorities. Specific cost-effectiveness analyses should be performed.

Guideline development team

Coordinator: Basilio Pintaudi, diabetologist.

Panel members: Daniela Brutomesso, diabetologist;Angela Grelli, diabetologist; Liliana Indelicato, psycholo-gist; Edoardo Mannucci, diabetologist; Basilio Pintaudi,diabetologist; Andrea Pizzini, general practitioner; Riccardo Schiaffini, pediatric diabetologist; Federico Spandonaro,economist; Katja Speese, nurse; Rita Stara, representative of people with diabetes; Giovanni Targher, diabetologist;Marilena Vitale, diettian; Riccardo Candido, diabetologist.

Evidence Review Team: Valentina Anelli, diabetolo-gist; Elisabet Romeo, diabetologist; Antonio Migliore1,biomedical engineer; Massimiliano Orso1, epidemiolo-gist; Daniela D'Angela1, economist; Barbara; Barbara Polisten1,economist.

${}^{1}$ C.R.E.A Sanita, Centro per la Ricerca Economica Appli-cata in Sanita

External reviewers: Domenico Cucinotta ${ }^{1}$ , Francesco Dotta ${ }^{2}$ , Dario Iafusco ${ }^{3}$

${}^{1}$ University of Messina; ${}^{2}$ University of Siena; ${}^{3}$ Univer-sity of Naples

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00592-025-02569-1.

Funding No external financial support was collected for the develop-ment of this guideline. Travel expenses for panel meeting were paid for by Societa Italiana di Diabetologia. Members of Panel did not receive any payment for their work in developing the guideline.

Declarations

Conflict of interest The assessment of interests of members of the Guideline development team is aimed at determining conflicts of in-terest for each question and the actions needed for their management in the process of elaboration of the Guideline. The assessment is based on the policy of the Istituto Superiore di Sanita for the management of conflicts of interest in the development of Guideline. Each interest is assessed for its nature, type, relevance for the content of the Guideline,economic value, timing and duration. The assessment includes the fol-lowing information which can be of help in determining the extent to which the competing interest could reasonably affect the expert's position: type of interest; relevance for the content of the guideline;timing and duration; position of the expert in the organization(in case of institutional interests). With respect to type of potentially competing interests, these include:(1) Economic interests, i.e. financial relation-ships with organizations directly producing goods or services relevant for the guideline topic. Economic interests include any monetary trans-action or value related to payments for services, property shares, stock options, patents and royalties. Relevant interest can be personal, relat-ed to family members or institutional(i.e. related to the organization in which the expert works).(2) Indirect interests, such as career advance-

ment, social position and personal beliefs. Interests considered can be:1. Economic interests, i.e. financial relationships with organizations involved in products or services relevant for the subject of the guide-line, including any direct payment for services, property shares, stock options, and patents or copyright royalties). Economic interests can be either: (a) personal economic interest, i.e. related to a personal finan-cial benefit; (b) familial economic interest, i.e. related to the income of family members; (c) institutional economic interests, i.e. related to benefits for the institution in which the subject works. 2. Intellectual interests, i.e. benefits for career advancement and social status. Both economic and intellectual interests can be specific (i.e. directly related to the subject of the guideline) or non-specific (when they are not re-lated to the content of the guideline).

Any reported potentially conflicting interest is classified as: Level 1(minimal or not relevant): no action needed. Level 2(potentially rele-vant): this can be managed either with full participation to the develop-ment of the guideline with public disclosure of the conflict of interest at the end of the recommendation related to the interest; exclusion of the subject with the competing interest form the discussion of those recommendations possibly influenced by the competing interest. Level 3(relevant): this can be managed with the exclusion of the subject with the competing interest from the discussion of possibly affected rec-ommendation, or with the total exclusion of the subject with compet-ing interest from the elaboration of the guideline. All members of the panel and of the evidence review team filled out annually a declaration of potential conflicts of interest, which were collectively discussed to determine their relevance. In all cases, the reported conflicts were con-sidered minimal or not relevant(Level 1); therefore, all components of the panel and of the evidence review team participated to the elaborata-tion of all recommendations.

Panel members: Edoardo Mannucci received fees for training activi-ties from Mundipharma and speaking fees from Abbott, Eli Lilly e Novo Nordisk; Riccardo Candido received consulting fees from Boeh-rlinger Ingelheim, Eli Lilly, Merck, Menarini and Roche, and speaking fees from Abbott, Eli Lilly, Mundipharma, Novo Nordisk and Sanofi;Basilio Pintauid received consulting and/or speaking fees from Eli Lil-ly, Novo Nordisk, Theras and Pikadare; Giovanni Targher received con-sulting fees from Novartas; Daniela Bruttomesso received consulting and/or speaking fees from Riccardo Schiaffini received consulting and/or speaking fees from Abbott, Sanofi; Federico Spandonaro received consulting and/or speaking fees from Eli Lilly, Janssen, Sanofi, Servi-er, Takeda, Allergan, Amgen, Astellas, Boehringer Ingelheim, Gilead,Gore, Janssen, Lundbeck, Nestlé, Novartis, Novo Nordisk, Sanofi; the others Members have no interest to declare.

Evidence review team members: Daniela D'Angela and Barbara Polistena received consulting and/or speaking fees from Accord HCI,Eli Lilly,Allergan,Amgen, Astellas,Bayer,Biogen,Boehringer In-gelheim,Chiesi, Daiichi Sankyo, Fond. MSD, Ipsen, Janssen, Lund-beck,Nestlé,Novartis,Novo Nordisk,Sanofi,and Dario Members have no interest to declare.

External reviewers: Domenico Cucinotta, Francesco Dotta, and Dario Iafusco have no interest to declare.

References

1. Magliano DJ, Boyko EJ, IDF Diabetes Atlas 10th edition sci-entific committee(2021) IDF Diabetes Atlas, 10th edn. Interna-tional Diabetes Federation, Brussels

2.https://www.epicentro.iss.it/diabete/epidemiologia-italia.Accessed 10 Aug 2024

Acta Diabetologica

3. Russo G, De Cosmo S, Di Bartolo P et al (2024) The quality of care in type 1 and type 2 diabetes-a 2023 update of the AMD annals initiative. Diabetes Res Clin Pract 213:111743

4. CINECA, Osservatorio ARNO Diabete 2017 http://arno.cineca.it/journal/wp-content/uploads/2017/10/Diabete_2017_con-coper tina.pdf. Accessed 10 Aug 2024

5.Guyatt GH, Oxman AD, Vist GE et al(2008) GRADE: an emerg-ing consensus on rating quality of evidence and strength of rec-ommendations. BMJ 336:924-926

6.Alonso-Coello P,Schinemann HJ,Moberg J et al(2016)GRADE evidence to decision(EtD) frameworks: a systematic and trans-parent approach to making well informed healthcare choices. 1:Introduction. BMJ(Clinical Research Ed) 353:i2016

Publisher's Note Springer Nature remains neutral with regard to juris-dictional claims in published maps and institutional affiliations.