“Mellitus Metrics”—Systematic Review of Glucometric Reporting within Hospital-Based Diabetes Studies (2006-2023)

Abstract

Background:

Suboptimal inpatient glycemia is associated with adverse outcomes, including infection, length of stay, and hospitalization costs. Interventions to improve inpatient glycemia may benefit from standardization of in-hospital glycemic measurement and reporting.

“Glucometrics,” as coined by Goldberg et al (2006), proposes models and metrics that allow quantitative inpatient glycemic data analysis. This systematic review investigates the actual use of “glucometric” terminology and its derivations since conception.

Methods:

Original research articles on “glucometrics” and its derivations in inpatient contexts, published between 2006 and 2023, were searched in five databases. Studies were screened and extracted through PRISMA-compliant review software (Covidence®) and systematically reviewed.

Results:

Of 767 studies identified, 44 were included for final review. Study settings included non-critical care wards (n=19), critical care (n=6), and both (n=13). Of the Goldberg models, “patient-day” was most used (n=33). Most studies (n=30) referred to “glucometrics” per the original description. An increase in the introduction of new metrics (e.g., time-weighted averages, adverse glycemic days, and glucose excursions) was seen over the study period, as well as an increase in the use of “glucometric” to refer to glycemic measurement/reporting in general.

Significant variation in thresholds defining hyperglycemia/hypoglycemia existed between studies, where hyperglycemia ranged between 140 and 432 mg/dL (most commonly 300 mg/dL), while the hypoglycemia ranged between 40 and 70 mg/dL (most commonly 70 mg/dL).

Conclusion:

This systematic review provides insights into contemporary use of glucometric terminology, highlighting the lack of consensus on a standardized approach toward analyzing inpatient glycemia, and the need for glucometric harmonization to improve inpatient glycemia and diabetes care.

Keywords

adverse glycemic day diabetes glucometrics hospital inpatient patient day mean glucose

Introduction

In the United States, approximately a quarter of hospitalized adults have diabetes, with prevalence rising by 2.5% annually between 2000 and 2018.¹ Blood glucose (BG) abnormalities increase risks of infection, mortality, and longer hospital stays, driving up costs.^2-5 Certain groups, such as those with acute coronary syndrome, strokes, community-acquired pneumonia, or undergoing surgery, are at higher risk of hyperglycemia.⁶ Diabetes and hyperglycemia prolong hospital stays and increase costs.^7,8 Effective hospital glucose management improves clinical outcomes, and reduces diabetes-related morbidity and costs to health care systems.^8,9

Optimizing inpatient diabetes care requires effective methods for monitoring and reporting glycemia. While outpatient monitoring with hemoglobin A1c (HbA1c) and continuous glucose monitoring (CGM) is well-established,^10-12 no universal metric exists for assessing inpatient glycemic status during short hospital stays. This limits quality assessments, evaluation of inpatient quality improvement initiatives, and benchmarking between institutions.^2-4,11,13-16

In 2006, Goldberg et al proposed “glucometrics,” which refers to standardized methods of analyzing inpatient blood glucose data and tracking glycemia over time.¹³ Such methods, once validated and widely used, allow objective comparisons between institutions, as well as within institutions over time.¹³

For glucometric comparisons, Goldberg et al proposed three distinct units of analysis pertaining to mean glucose metrics, each with advantages and disadvantages. The first is by “population,” where all BG levels of an inpatient population are grouped, but heavily influenced by the frequent BG testing in cases with hyperglycemia or hypoglycemia. The second, the “patient-stay” model, aggregates BG results for individual patients over their hospital stay into a single mean value. Lastly, the “patient-day” model groups BG results by each patient and for each calendar day of the admission, minimizing the impact of length of stay and testing frequency.

The potential of glucometrics has been extensively explored, and its meaning has expanded since the original description, especially with increase in evidence for the negative effects of in-hospital adverse glycemia. The increased use of terminology in different contexts and the variability in meanings complicate efforts for more widespread use of glucometrics in the hospital setting.

The primary aim of this systematic review was to evaluate how “glucometric” concepts and terminology have been used in the published literature between 2006 and 2023 compared to the original description by Goldberg et al.

Methods

This qualitative systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines.¹⁷ The review protocol was prospectively registered (https://doi.org/10.17605/OSF.IO/5TNMR). The PRISMA checklist is provided in the Supplementary Materials (Supplementary 1).

Eligibility Criteria

This review included studies applying glucometric concepts in the inpatient setting. The types of studies included were original research articles, including interventional and observational studies. Reviews, perspective articles, conference abstracts or other non-experimental papers were excluded. The search was limited to papers published in English, with full text articles obtained through institutional access methods.

Information Sources and Search Strategy

To identify eligible studies, we performed a search of five electronic databases (PubMed, OVID, Medline, Web of Science, and Cochrane Library) from January 1, 2006, to August 15, 2023. Free-text keyword search terms were combined with the Boolean operators “OR” and “AND.” The final search strategy used was: (inpatient OR in-patient OR ward) AND (glucometric OR glucometrics OR gluco-metric OR gluco-metrics OR “patient-day mean glucose” OR “patient-days mean glucose” OR “patient-stay mean glucose” OR “patient-stays mean glucose” OR “patient-day-weighted mean glucose” OR “mean glucose” OR “adverse glycaemic day” OR “adverse glycemic day”). The detailed search strategy is presented in the Supplementary Materials (Supplementary 2).

Selection Process

Two investigators (JWSY and RW) independently reviewed studies using PRISMA-compliant eligibility forms in Covidence software (Veritas Health Innovation, Melbourne, Australia). Any discordance was resolved through seeking consensus, involving a third investigator (RDB) when necessary. Duplicate records were removed, and the remaining study titles and abstracts were screened before retrieving and reviewing full texts to identify eligible studies.

Data Extraction

Data were extracted independently in Covidence software by two investigators (JWSY and RW), who then verified accuracy through simultaneous comparisons. Extracted data included article title, first author, publication year, full journal title, study country, study design, study population of interest, study purpose, total number of patients, glucometric terms used, usage of “glucometric,” glucose testing methods, brand of glucose meter or analyzer, and thresholds for hyperglycemia and hypoglycemia.

Data Synthesis

Extracted data were assessed for the frequency of glucometric terminology use over time and its use compared to the original description by Goldberg et al. Additionally, glucose measurement methods and threshold criteria for defining hyperglycemia and hypoglycemia were assessed.

Results

A total of 767 studies were identified, of which 44 studies were included in the final review (summarized in Table 1).^{3,4,8-11,13,18-54} The PRISMA flow diagram detailing the study selection process is depicted in Figure 1. A significant number of studies were excluded due to duplicate records, conference abstracts, or being conducted in outpatient settings.

Table 1.

Summary of key features for included studies.

Reference number, First author, Year, Location,Study design	Study purpose	Study pop. of interest	Pt. pop. size	Manner of use of the word “glucometric”(Refer to table caption for reference to 1-6)	Glucometric terms used as coined by Goldberg et al. (2006)	Additional glucometry terms/ models not in Goldberg et al. (2006)	Glucose sample method, analyzer	Hyperglycemia BG threshold	Hypo-glycemiaBG threshold
3, Kyi, 2019, Australia, Prospective observational study (cross-sectional)	To assess glucometric outcomes and to estimate the incidence of hypo- and hyperglycemia among non-critical care inpatients in a major Australian hospital	Non-critical care inpatients with diabetes or newly detected hyperglycemia admitted to medical and surgical wards	441	2, 5, 8	Glucometric, Population model, Patient-stay model, Patient-day model, Mean, Any hypereor hypoglycemia	“Adverse glycemic days.” (patient-days with any BG <4 or >15 mmol/L)	Capillary fingerprick, StatStrip (Nova Biomedical, MA, USA) via local wireless network	>10 > 13.9 >15 > 16.7 mmol/L	<4 mmol/L Severe: <3 mmol/L
4, Maynard, 2014, United States, Cross-sectional study	To build a comprehensive web-based glucometric data reporting software (to allow institutions to assess their baseline glucometrics and the impact of their improvement efforts, and allow for comparison of their performance with others)	Inpatients with at least four BG readings over a minimum two consecutive days	Unclear (59 995 ICU admissions) (non-ICU admissions not stated)	2, 4	Glucometrics, Patient-day, Patient-stay		Capillary fingerprick and blood gas and venous plasma, not stated for all	Severe: >299 mg/dL	<70<40 mg/dL
8, Barmanray,2022,Australia,Cross-sectional study	To establish value of longitudinal institutional glucometric benchmarking to assess impact of quality improvement changes	Inpatients with diagnosis of diabetes or at least one BG ≥11.1 mmol/LNon-critical care medical and surgical wards	588	2, 4, 8	Glucometric, Patient-day mean glucose, Patient-stay mean glucose, Population mean glucose, Mean level with any BG in target levels (hypo- or hyperglycemia) for each patient-day, patient-stay, population model	“Adverse glycemic day”	Capillary fingerprick, StatStrip (Nova Biomedical, MA, USA) via local wireless network	>10 >13.9> 15 >16.7 mmol/LNote:USA bench-marks: >10, >13.9, >16.7 mmol/L	<3.9<2.8<2.2 mmol/LNote:USA bench-marks: <3.9, <2.8, <2.2 mmol/L
9, Kyi,2019,Australia,Randomized controlled trial	To investigate if early electronic identification and bedside management (proactive care) of inpatients with diabetes/ hyperglycemia improves glycemic control and clinical outcomes (HAI) in non-critical care	Non-critical care adult inpatients with diabetes or hyperglycemia	1002	2, 7	Glucometric, Patient-day mean glucose, Patient-day	“Adverse glycemic days”, “Good diabetes days:	Capillary fingerprick, StatStrip (Nova Biomedical, MA, USA)	>10 or >15 mmol/L	<4 or <3 mmol/L
10, Saulnier,2022,United States,Cross-sectional study	To develop a spatial representation/mapping of hospital wards to assess variation in inpatient glucose control	Non-critical care, inpatients with diabetes from 20 medical and surgical units	2344	2	Glucometric, Patient day-weighted mean, % of patient hospital days with hypoglycemia		Capillary fingerprick, StatStrip (Nova Biomedical, MA, USA)	Not stated	<70 mg/dL
11, Thomas,2008,United States,Qualitative/Descriptive study	To describe process of building an open access Web application capable of computing glucometrics on uploaded data	Inpatients	Not stated	2, 3	Glucometrics, Patient-stay, Patient-day, Patient-sample (population model)		Not stated, not stated	>299 mg/dL	<70 mg/dL
13, Goldberg,2006,United States,Cross-sectional study	To propose standardized metrics for inpatient glycemic control to be developed and validated	General medical hospital ward	118	2, 5, 7, 8	Glucometrics, Patient-day, Patient, Population, Mean BG, Median BG, % BG in range (80-139 mg/dL), % hypo- or hyper-glycemic events by Patient-day, Patient-stay, and Population models		Capillary fingerprick, SureStep Hospital (LifeScan, CA, USA)	>300 mg/dL	<60 mg/dL
18, Maynard,2017,United States,Pre vs post intervention observational study	To assess glycemic impact of quality improvement interventions (using collaborative interventions which included standardized insulin infusion protocols, hypoglycemia prevention bundles, audit and feedback, education, and “measure-vention”)	Critical care (ICU) inpatients with at least four POC-BG readings over at least two calendar daysExclusion criteria: non-critical care patients, outpatients, pediatric patients, and maternal child health patients	10 112	2	Glucometric, Patient-day model, Day-weighted mean BG	“Glucometric summaries”	Capillary fingerprick, not stated	Severe: >299 mg/dL	<70 mg/dL
19, Maynard,2017,United States,Pre vs post intervention observational study	To assess glycemic impact of quality improvement interventions to standardize insulin management and glucometric reporting	Acute care (non-ICU), inpatients admitted to medical, surgical, orthopedics, or telemetry units with at least four POC-BG readings in at least two calendar days of an inpatient stayExclusion criteria: critical care patients, outpatients, pediatric patients, and maternal child health patients	62 431	2	Glucometrics, Patient day-weighted mean blood glucose, Patient-day model	Society of Hospital Medicine glucometric bench marking document	Capillary fingerprick, not stated	Severe: >299 mg/dL	<70 mg/dL
20, Mclver,2009,Australia,Pre vs post intervention observational study	To assess glycemic impact of improvements in insulin management using new standardized forms for intravenous and subcutaneous insulin prescription and blood glucose management	Non-critical care adult inpatients in four Queensland Health tertiary and regional hospitalsExclusion criteria: pediatrics, mental health, ICU	199	2, 3(Referring to Yale School of Medicine webpage called “Gluco-metrics”)	Glucometrics, Percentage of patient days with any hypoglycemic event or hyperglycemic events		Not stated, not stated	>16.6mmol/L	<3.9mmol/L
21, Cheung,2011,Australia,Pre vs postintervention observational study	To assess the glycemic effect of implementing a new subcutaneous insulin prescription chart	All patients with diabetes who received subcutaneous insulin orders	205	2	Glucometrics, Patient-day, Patient-day mean glucose, Whole population (population model), Individual patient (patient-stay model)		Capillary fingerprick, Optium (Abbott Diabetes Care, CA, USA)	≥ 10 or≥ 16mmol/L	< 4 mmol/L
22, Bogun,2022,United States,Pre vs postintervention observational study	To assess glycemic and clinical outcomes in patients undergoing CABG, after the implementation of case-based diabetes workshops for health care providers	Patients who underwent isolated CABGExclusion criteria: patients who underwent CABG concomitantly with valve or aortic procedures	1970	“Gluco-metric” word not used, but “glucose metrics” used instead	Glucose metrics, Mean patient day glucose, Patient-day, Percentage patient days with mean glucose in target range, Patient-stay		Capillary fingerprick, not stated	Severe: >250 mg/dL	Severe: < 50 mg/dL
23, Lleva,2014,United States,Pre vs post intervention observational study	To determine if the prevalence of hyperglycemia in US ICUs had increased after the NICE-SUGAR study’s results were reported	ICU	408 790 patient-days	2, 3(Referring to Yale Glucometrics website name)	Glucometrics, Patient sample (population model), Patient-day, Median patient-day glucose, Patient-days with mean glucose 110-179 mg/dL, Patient-stay		Capillary fingerprick, not stated	≥ 180 mg/dLSevere:≥300 mg/dL	<70 mg/dL or <40 mg/dL
24, Shelden,2021,United States,Pre vs post intervention observational study	To integrate a previously validated IV insulin calculator into electronic health record (Epic) and instituted it in the cardiovascular ICU, then evaluate the calculator’s glycemic impact	Cardiovascular ICU nurses	79registered nurses	1	Glucometrics		Not stated, not stated	>250 mg/dL	<70 mg/dL<54 mg/dL<40 mg/dL
25, Shabbir,2010,United States,Pre vs post intervention observational study	To assess glycemic impact of implementation of a real-time nursing intervention	Adult, non-critically ill, non-obstetric patients	653	2	Glucometrics, Patient day-weighted mean glucose		Capillary fingerprick & venous plasma, Accu-Chek (Roche, Switzerland) & Siemens RXL MAX (Siemens, IL, USA)	>180 mg/dL	Severe: <50 mg/dL
26, Bansal,2015,India,Pre vs post intervention observational study	Quality improvement intervention: introduction of a standardized insulin infusion protocol in cardiac surgery	ICU,Post-cardiac surgery cohort	1675	4(No reference to Goldberg et al. (2006))	Glucometrics, Mean BG (population model)		Blood gas, ABL800FLEX	>250 mg/dL	<60 mg/dL
27, Maynard,2009,United States,Pre vs post-intervention observational study	To assess the glycemic impact of interventions of insulin use patterns, glucose monitoring	All adult inpatients with diabetes or hyperglycemia on non-critical care units from November 2002 through December 2005Exclusion criteria: those who did not have either a discharge diagnosis of Diabetes or demonstrated hyperglycemia	9314	2, 4(Refers to the Society of Hospital Medicine Glycemic Control Task Force summary)	Glucometrics, Patient-stay, Patient-stays with hypoglycemia, Patient-day, Day-weighted mean glucose, Patient-days with hypoglycemia, Day-weighted median, Patient-day median glucose		Capillary fingerprick, not stated	≥180 mg/dL	≤60 mg/dLSevere: ≤40 mg/dL
28, Kao,2020,Singapore,Pre vs post intervention observational study	To assess glycemia impact of the multi-faceted intervention program	Non-critical care wards, patients with prescription of DM medications or any capillary BG >11.1mmol/L during admissionExclusion criteria: admissions shorter than 24 hours, patients with age <16, Pediatric and labor wards, inconsistent data	14 657	2	Glucometric, Patient-days with hypoglycemia / hyperglycemia, Patient-day mean glucose, Patient admissions affected by hypoglycemia / hyperglycemia, Patient admission level mean glucose (patient-stay model)		Capillary fingerprick, not stated	> 14 mmol/LSevere: >20 mmol/L	< 4, <3,<2.5 mmol/L
29, Brand,2019,United States,Pre vs post intervention observational study	Assessing the glycemic impact of quality improvement intervention: a basal-bolus insulin therapy protocol in non-critical care settings	Non-critical care,Non-pregnant adults with T2DM admitted to general medical or surgical inpatient unit and hospitalized for at least two days	40 391	2, 4,	Glucometric, Proportion of days during a patient’s hospital stay (patient-day model), Mean BGL (population model)(Used “patient-day” model as per Goldberg model but without using PDMG)		Capillary fingerprick, not stated	>200 mg/dL>300 mg/dL	<70 mg/dL <40 mg/dL
30, Mendez,2015,United States,Retrospective pre vs post intervention observational study	To assess glycemic impact of the DINGS program (automated EMR-based glucose alert system) on glycemic control and the prevalence of hypoglycemic events in patients with diabetes and hyperglycemia	Admissions with glucose values available within two years prior to 19 months after establishment of the “DINGS” process (January 2008 to September 2013)	7133	2	Glucose metrics, Glucometrics, Patient-day, Patient-day mean glucose, Patient-days with any hyperglycemia, % of patient-days with mean glucose in range of 90-180 mg/dL, % of patient-days with any hypo- or hyperglycemia”		Capillary fingerprick and venous plasma, Accu-Chek Inform (Roche, Switzerland) and not stated	>180 mg/dLSevere: >350 mg/dL	<70 mg/dLSevere: <50 mg/dL
31, Engle,2016,United States,Pre vs post intervention observational study	Quality improvement project to redesign a hospital meal delivery process to shorten the time between pre-meal blood glucose monitoring and corresponding insulin administration	Patients receiving mealtime insulin, with or without diabetes diagnosis	Not stated	2, 4	Glucometrics, Patient-day, % patient-days with specific blood glucose level		Capillary fingerprick, not stated	>300 mg/dL	<70 mg/dL
32, Goswami,2019,United States,Pre vs post intervention observational study	To assess the impact of multidisciplinary process improvement interventions on glycemic control in the inpatient setting	Adult non-critical care medical ward inpatients	8943	2, 8	Glucometric, Mean of patient daily glucose values (patient-day model), Mean of patient glucose values (patient-stay), Analyses of frequency of hyper- or hypoglycemia at each site (population model)	“Time-weighted mean of patient daily glucose values” (area under glucose-time curve)“Time-weighted mean of patient glucose values” (area under glucose-time curve)	Capillary fingerprick, not stated	>200 mg/dLExtreme: >299 mg/dL	<70 mg/dLExtreme: <40 mg/dL
33, Mulla,2015,United States,Pre vs post intervention observational study	To explore the relationship between education for inpatient diabetes providers and the utilization of insulin order sets, inpatient glucometrics, and length of stay in a large health care system.	Critically ill and non-critically ill inpatients available for review annually between 2008 and 2011,included those with discharge diagnosis of diabetes that were retrospectively identified	22 000	4(No reference to Goldberg models)	Glucometrics		Capillary fingerprick and venous plasma, not stated	>180 mg/dL	<70 mg/dL
34, Weinberg,2010,United States,Cross-sectional study	To validate a method to manage frequently repeated glucose measurements, in order to accurately report rates of hypo/hyperglycemia	Non-ICU,Adult medical and surgical wards in August of 2006	786	2	Glucometrics, Patient-day model		Capillary fingerprick, not stated	>350 mg/dL	<60 mg/dL
35, Koziol,2015,United States,Cross-sectional study	To introduce a simple metric of glucose variability: glucose excursion rate.	All patients admitted between 2009 and 2011 and who underwent POC glucose testing over at least a 24-hour period	2024	4, 7, 8	Glucometric, Day-weighted mean glucose, Stay-weighted mean glucose	“Glucose excursion”	Capillary fingerprick, SureStep Hospital (LifeScan, CA, USA)	>180 mg/dL	<70 mg/dL
36, Kosiborod,2008,United States,Cohort study	To compare different glucose metrics incorporated into glucose assessments during hospitalizations vs admission glucose alone as predictors for post-AMI in-hospital mortality	All patients hospitalized with the primary discharge diagnosis of AMI who had at least one glucose measurement during the first 24 hours after admission and at least one documented abnormal troponin I or T or creatinine kinase-MB fracture	16 871	2, 4, 5	Glucose metrics, Mean glucose, Hospitalization mean glucose (patient-stay model)	“Time-averaged glucose”, “Hyper-glycemic index”	Capillary fingerprick and venous plasma, not stated	≥ 200 mg/dL	<110 mg/dL
37, Buchs,2010,Israel,Cross-sectional study	To report on glucose control of all patients with diabetes hospitalized in general medicine wards	General medicine ward patients with either previously known or newly diagnosed diabetes mellitus	2475	3, 4	Glucometrics		Capillary fingerprick, Accu-Chek Inform II (Roche, Switzerland)	Reported on results using >200 mg/dLNote: typo in the body defining hyperglycemia as >300 mg/dL	< 60 mg/dL
38, Dharma, 2024 (published online ahead of print in 2023),Australia,Cross-sectional study	To assess the biochemical validity and clinical relevance of hospital coding-defined hypoglycemia HAC (hospital-acquired complication)	Cohort 1: inpatients discharged who had coding of hypoglycemiaCohort 2: inpatients with diabetes	395	4(No reference to Goldberg et al. (2006))	Glucometry, Glucometric, Patient-stay model		Capillary fingerprick and blood gas and venous plasma, not stated for all	Not stated	“Level 1”: 3.0-3.9 mmol/L”“Level 2”: <3 mmol/L”“Level 3”: severe event (e.g. altered mental state, requiring urgent medical attention)
39, Guerra,2010,United States,Cross-sectional study	To evaluate the glycemic impact of implementing a computerized physician order entry (CPOE)-based hyperglycemia inpatient protocol (HIP)	Previous diagnosis of T1DM/T2DM, admission to medicine wards	438	2, 4(Gluco-metric analysis performed using Yale Center for Medical Informatics web-based application)	Glucometric, Patient-day glucose, Mean blood glucose level per patient-day, On-target glycemia rates per patient-day, hypoglycemia patient-day, hyperglycemia patient-day		Capillary fingerprick, not stated	>180 mg/dLSevere: >300 mg/dL	<70 mg/dLSevere: <50 mg/dL
40, Khatib,2014,India,Cross-sectional study	To describe glucose results assessed by glucose readings obtained through manual chart abstraction	All patients admitted to ICU	26	2	Glucometrics, % of values in a specific range (population model), Patient-stay, Average glucose (for both patient-day / stay), % of patient-stays / days with average BGL within target (140-180 mg/dL), % of patient-stays / days with any hypo- or hyperglycemia		Not stated, not stated	>300 mg/dL	<70 mg/dL
41, Bersoux,2014,United States,Retrospective cross-sectional observational study	To report data on glucose control from 635 US hospitals	Adult inpatients from 635 US hospitalsICU and non-ICU	2 612 966	2, 6	Patient-day-weighted mean glucose, Patient-days with (severe) hyper- or hypoglycemia		Capillary fingerprick, Accu-Chek Inform II (Roche, Switzerland)	>180, >200, >250, >300, >350, >400 mg/dL	<70,<60,< 50,<40 mg/dL
42, Saulnier,2017,United States,Cross-sectional study (Applying mathematical modeling)	To apply methods of damped trend analysis to forecast inpatient glycemic control	Non-ICU, Inpatients with diabetes mellitus	55 328 patient-day-weighted mean glucose	2	Glucometrics, Patient-day-weighted mean(The Goldberg model of PDWMG was incorporated into linear modeling/Forecasting)		Capillary fingerprick, not stated	Not stated	Not stated
43, Bersoux,2013,United States,Retrospective cross-sectional observational study	To evaluate trends in glycemic control in US hospitals over two years	ICU and non-ICU,Adult inpatients from 126 US hospitals	1 666 911	6	Patient-day-weighted mean glucose, Patient-days with at least one hypoglycemia		Capillary fingerprick, not stated	>180, >200, >250, >300, >350, >400 mg/dL	<70 mg/dL or <40 mg/dL
44, Rao,2021,United States,Cohort study	To assess glycemic impact of an ICU computerized insulin protocol (GENIE)1. GENIE (computerized IIT) vs FORMULA (traditional IIT)2. GENIE vs euglycemic no-insulin cohort	Surgical ICU, post open heart surgery requiring intensive insulin therapy (IIT)Criteria for IIT:1) pre-existing diabetes OR stress hyperglycemia (2 successive BGs > 140 mg/dL)2) an order for IIT3) an insulin infusion dispensed by pharmacy and administered for ≥ 12 h4) repetitive BG monitoring for ≥12 h	1404	1	Glucometrics	“Time-averaged glucose (TAG)”, “Time-weighted exposure to glucose(mg/dL/h)”	Not stated, not stated	>140 mg/dL	<70 mg/dLSevere: <50 mg/dL
45, Munshi,2020,United States,Retrospective cohort study	To compare PTDM incidence and risk factors between kidney and liver transplant recipients	Patients who underwent liver-only and kidney transplant patients	753	2	Glucometrics, Patient-stay mean glucose		Capillary fingerprick, Accu-Chek (Roche, Switzerland)	Fasting glucose: >126 mg/dL	Not stated
46, Schnipper,2006,United States,Prospective observational cohort study	To determine the current state of glucose management on an academic hospitalist service and the relationship between insulin-ordering practices and glycemic control	Patients admitted to three General Medicine Service teams with either a known diagnosis of diabetes or inpatient hyperglycemia (random glucose > 200 mg/dL)Exclusion criteria: admission for DKA, HHS, gestational diabetes	107	2, 6	Mean glucose level per patient-day		Capillary fingerprick, not stated	>180 mg/dL	<60 mg/dL
47, Munoz,2012,United States,Retrospective cohort study	To assess glycemic impact of implementing hospital-wide inpatient glucose management quality improvement program	Adult inpatients with diagnosis of diabetes or hyperglycemia (two POC BGLs > 180mg/dL or ICD-i codes for impaired fasting glucose), with ≥ 5 POCT BGL readings during their hospital stay	25 160	2, 4	Glucometrics, Patient day-weighted mean blood glucose, Patient-day, % of patient-days with hypo- or hyperglycemia		Capillary fingerprick, not stated	>180 mg/dLExtreme: >250 mg/dL	<70 mg/dL
48, Munshi,2020,United States,Retrospective cohort study	To describe glycemic trends and PTDM incidence after heart transplant	First time heart-only transplant between 2010 and 2015	152	4	Glucometrics, mean patient hospital stay glucose (patient-stay)		Capillary fingerprick and venous plasma, Accu-Chek Inform (Roche, Switzerland) and not stated	fasting glucose >126 mg/dL	Not stated
49, Seheult,2015,Ireland,Qualitative/Descriptive study	To develop an automated system to process POC testing blood glucose results stored on network server and to present the daily analysis in a user-friendly interface to the diabetes consult service as a glucose alert system	All inpatient POC test blood glucose resultsExclusion criteria: pediatric wards, day wards, outpatient departments, catheter laboratory, surgical theaters, dialysis department, emergency department, acute medical unit	Not stated	1, 3	Glucometry, Glucometrics		Capillary fingerprick, Accu-Chek Inform II (Roche, Switzerland) via local wireless network	Mild:10-15 mmol/LModerate: 15-20 mmol/LSevere: >20 mmol/L	<4 mmol/L
50, Chen,2021,Singapore,Qualitative/Descriptive study (Creation of a software tool to calculate glucometric measures)	To facilitate monitoring of the quality of inpatient glycemic control by describing process of developing open-source software tool that computes glucometrics	Not stated	Not stated	2	Glucometrics, Patient-stays, Patient-days, Patient-samples (population model)	“Hyper-glycemia Index AUC”	Capillary fingerprick, Accu-Chek Inform (Roche, Switzerland)	≥ 14, ≥ 20, ≥ 24 mmol/LHyperglycemia Index AUC >10 mmol/L/LOS in hours	<4, <3, <2.5 mmol/L
51, Rushakoff,2014,United States,Qualitative research	To describe qualitative outcomesglycemic control program through the use of an external mentoring program	hospital-based interdisciplinary teams to advance inpatient glucose management	139 health professionals	4(Refers to hospital sites that collects and reports glucose data, but no quantitative results provided)	Glucometrics		Not stated, not stated	Not stated	Not stated
52, Schnipper,2010,United States,Randomized controlled trial	To determine the effects of a CPOE (computer provider order entry) vs usual care, upon the inpatient management of diabetes and hyperglycemia	Consecutive patients admitted to general medicine with either known diabetes mellitus or inpatient hyperglycemiaExclusion criteria: pregnancy, HHS, DKA, receiving TPN, had another indication for IV insulin, palliative care, zero POC-BG	179	2, 6	Patient day-weighted mean glucose, Patient-day		Capillary fingerprick, not stated	>300 mg/dL	<60 mg/dLSevere: <40 mg/dL
53, Bernard,2011,United States,Randomized controlled trial	To examine the impact of implementing a formal subcutaneous insulin protocol (i.e. aspart insulin protocol) for treatment of hyperglycemia in the ED, coupled with rapid initiation of detemir-aspart insulin protocol for patients admitted to the hospital	Presenting to ED with a history of T2DM for at least three months & initial POC-BG ≥ 200 mg/dL, Ages 18-80, prior therapy with dietary management/ oral agents/ insulinExclusion criteria: DKA, HHS, ICU admission, surgical intervention, positive pregnancy test, inability to give informed consent (e.g. acute drug/ alcohol intoxication, active mental illness), clinically significant liver disease (e.g. ALT or AST > 3 times the upper limit of normal), end-stage renal disease requiring dialysis, T1DM	176	2, 6	Patient day-weighted mean glucose, Patient-days with hypoglycemia, Patient-stays with hypoglycemia		Capillary fingerprick, not stated	Did not report upon hyperglycemiaNote:Intervention protocol utilized BG >200 mg/dL as threshold	Moderate:50-69 mg/dLSevere:<50 mg/dL, <60 mg/dL
54, Lin,2015,Taiwan,Case–control study	To evaluate glycemic impact of an automated computerized platform to alert primary care teams and diabetes consultation teams to recurrent hyperglycemia (two successive POC BG ≥ 250 mg/dL on same day) on medical ward inpatients	Medical ward inpatients with diabetes or hyperglycemia	1644	2	Patient-day, Patient day-weighted mean glucose, Percentage patient day-weighted glucose ≥180 mg/dL, Patient-days with hypoglycemia (<70 mg/dL)		Capillary fingerprick, StatStrip (Nova Biomedical, MA, USA) via local wireless network	≥180 mg/dL	<70 mg/dL

1—As a general term for reporting inpatient glycemia.

2—Reporting on a Goldberg glucometric as a study outcome.

3—Proper Noun-Term used as the title of the software program.

4—Used to describe the glucose monitoring process/glucose measurement analysis.

5—Comparing differing glucometric models of both Goldberg and non-Goldberg models.

6—“Glucometric” term is not used.

7—Validating a glucometric measure.

8—Used non-Goldberg performance metrics.

Figure 1.

PRISMA flow diagram.

Study Characteristics

Of the 44 studies included, there were 16 pre- vs post-intervention observational studies,^18-3314 cross-sectional studies,^{3,4,8,10,13,34,35,37-43}six cohort studies,^36,44-48four qualitative studies,^11,49-51three randomized controlled trials,^9,52,53 and one case–control study.⁵⁴ The studies were primarily from North America (n=31),^{4,10,11,13,18,19,22-25,27,29-36,39,41-48,51-53} with 11 from the Asia-Pacific region^{3,8,9,20,21,26,28,38,40,50,54} one each from Israel³⁷ and Ireland.⁴⁹ Regarding hospital settings, 22 studies focused on non-critical care wards,^{3,8-10,13,18,19,22,25,27-29,32,34,37,39,42,45,46,48,52,54} 14 on both critical and non-critical wards,^{4,11,21,30,31,33,35,36,38,41,43,47,49,51} seven on critical care wards,^{18,23,24,26,40,44,53} with one unspecified.⁵⁰ The yearly distribution of studies is shown in Figure 2.

Figure 2.

Number of included studies by publication year.

Patient population size ranged from 26 to 2 612 966, with 11 studies including <500 patients,^{3,13,20,21,38-40,45,46,52,53} four studies between 500 and 1000 patients,^8,25,34,45 11 studies 1000-10 000 patients,^{9,10,22,26,27,30,32,35,37,44,54} eight studies 10 000-100 000 patients,^{18,19,28,29,33,36,42,47} and two studies with >100 000 patients.^41,43 Two studies reported the patient population size in “patient-days,”^23,42 two studies did not report any patient population size as they were not the target population of the study,^24,51 and four studies did not report patient population size,^11,31,49,50 even though they reported glucometric data from the target population.

Glucometric Utilization and Reporting

Glucometric models

Of the three glucometric models described by Goldberg et al,¹³ the “patient-day” model was utilized in 33 studies,^{3,4,8-11,13,18-23,25,27-32,34,35,39-43,46,47,50,52-54} the “patient-stay” model in 19 studies,^{3,4,8,11,13,21-23,27,28,32,35,36,38,40,45,48,50,53} and the “population” model in 11 studies (Figure 3(a)).^{3,8,11,13,21,23,26,29,32,40,50} A single glucometric model was used in 22 studies;^{9,10,18-20,25,26,30,31,34,36,38,39,41-43,45-48,52,54} two models in seven studies;^{4,22,27-29,35,53} and all three models in nine studies.^{3,8,11,13,21,23,32,40,50} Six studies did not use any models but simply used the term “glucometrics.”^{24,33,37,44,49,51} Figure 3(b) demonstrates comparable use of the three models in the first vs second halves of the time period assessed in this review.

Figure 3.

(a) Included studies, organized by the type of glucometric models used. (b) Included studies organized by models used between 2006-2014 and 2015-2023.

Glucometric reporting and metrics

Figure 4(a) depicts the variability in metrics being reported. “Mean glucose” was reported in 32 studies,^{3,8-10,13,18,19,21-23,25-30,32,35,36,39-43,45-48,50,52-54} “proportion of hyperglycemia/hypoglycemia” in 18 studies,^{3,8,10,13,20,28,30-32,34,38-41,43,47,53,54} while new metrics were reported in seven studies (“adverse glycemic days” [n=3];^3,8,9 “time-weighted average glucose” [n=3];^32,36,44 and “glucose excursion rate” [n=1]).³⁵ When comparing both halves of the examined time period (Figure 4(b)), new performance metrics not introduced by Goldberg et al increased in prevalence from 2015 to 2023.

Figure 4.

(a) Included studies, organized by performance metrics reported. (b) Included studies organized by metrics used between 2006-2014 and 2015-2023.

Studies’ main purpose of glucometric terminology usage

Some of the included studies utilized glucometric terminology in more than one way. The concept of glucometrics was utilized in 24 studies to assess intervention effectiveness;^{9,18-33,39,44,47,51-54} 12 studies to measure inpatient glycemia quality;^{3,8,34,36-38,40,41,43,45,46,48} and four as the subject matter in developing calculation software.^4,11,49,50 Eight studies proposed new performance metrics (“adverse glycemic days,” “time-weighted average glucose,” and “glucose excursion rates”).^{3,10,13,32,35,36,42,44}

Analytic modalities and devices

Capillary fingerprick glucose measures were utilized in 37 studies (Roche Accu-Chek in eight studies,^{25,30,37,41,45,48-50} Nova Statstrip in 5,^3,8-10,54 Lifescan SureStep in 2,^13,35 Abbott Optium in 1,²¹ and device not stated in the remaining 21)^{4,18,19,22,23,27-29,31-33,34,36,38,39,42,43,46,47,52,53}. Blood gas measures were used in three studies,^4,26,38 and plasma glucose measures in seven.^{4,25,30,33,36,38,48} Of the 44 studies, 31 reported a single glucose modality,^{3,8-10,13,18,19,21-23,26-29,31,32,34,35,37,39,41-43,45-47,49,50,52-54} seven reported multiple combinations of capillary point-of-care (POC), blood gas, and plasma glucose,^{4,25,30,33,36,38,48} while six did not specify the glucose modality used.^{11,20,24,40,44,51}

Glucometric Thresholds

Glucose thresholds used to define hyperglycemia (Table 2) and hypoglycemia (Table 3) varied between different studies.

Table 2.

Glucose thresholds used to define hyperglycemia.

Criteria		Number of studies
Any glucose measure exceeding threshold
(mg/dL)	(mmol/L)
140	7.8	1
180	10.0	11
200	11.1	5
250	13.9	6
270	15.0	3
288	16.0	1
300	16.7	14
350	19.4	3
360	20.0	3
400	22.2	1
432	24.0	1
Mean glucose exceeding threshold
(mg/dL)	(mmol/L)
180	10.0	9
200	11.1	2
250	13.9	4
270	15.0	2
300	16.7	3
350	19.4	1
400	22.2	1
Other
Fasting glucose 126 mg/dL (7.0 mmol/L)		2
Time-weighted average glucose 140 mg/dL (7.8 mmol/L)		1

Table 3.

Glucose thresholds used to define hypoglycemia.

Criteria		Number of studies
Any glucose measure below the threshold
(mg/dL)	(mmol/L)
70	3.9	30
60	3.3	9
54	3.0	7
50	2.8	9
45	2.5	2
40	2.2	10
Mean glucose below the threshold
(mg/dL)	(mmol/L)
110	6.1	1

Thresholds used to define hyperglycemia included any single BG measure above 140-432 mg/dL (7.8-24.0 mmol/L), or a mean glucose above 180-400 mg/dL (10.0-22.2 mmol/L). Rarely, studies defined hyperglycemia using fasting glucose or time-weighted mean glucose. The most commonly used hyperglycemia criteria were any glucose exceeding 300 mg/dL (16.7 mmol/L) (n=14)^{4,11,13,18-20,23,29,31,32,39-41,52} or 180 mg/dL (10.0 mmol/L) (n=11),^{3,8,21,30,33,35,39,41,46,49,54} or a mean glucose exceeding 180 mg/dL (10.0 mmol/L) (n=9).^{3,8,9,23,25,27,43,47,54} Multiple levels of hyperglycemia severity were reported in 17 studies,^{3,8,9,21,23,28-30,32,39,41,43,44,47,49,50,54} while a single threshold was used in 23.^{4,11,13,18-20,22,24-27,31,33,34-37,40,45,46,48,52,53}

Hypoglycemia was variably defined as any BG measure below 40-70 mg/dL (2.2-3.9 mmol/L). In one study, a mean glucose level was used to describe hypoglycemia.³⁶ The most commonly reported hypoglycemia criteria were any glucose measure below 70 mg/dL (3.9 mmol/L) (n=30)^{3,4,8-11,18-21,23,24,28-33,35,38-41,43,44,47,49,50,53,54} or 40 mg/dL (2.2 mmol/L) (n=10).^{4,8,23,24,27,29,32,41,43,52} Multiple levels of hypoglycemia severity were reported in 19 studies,^{3,4,8,9,23,24,27-30,32,38,39,41,43,44,50,52,53} while a single threshold was used in 21.^{10,11,13,18-22,25,26,31,33,34-37,40,46,47,49,54}

Discussion

This systematic review assessed the use of glucometric principles and terminology over time since 2006.¹³ Despite evolution in the use of glucometric as a term and as a concept over the two decades, we found that the majority of included studies still used “glucometrics” terminology consistent with the original description. Interestingly, direct reference to Goldberg’s study decreased over time, and more studies began to use “glucometrics” as a general concept. This suggests increasing familiarity and understanding of “glucometric” terminology in the literature, decreasing the need for direct reference to the original paper. There was also an increase in the introduction of new glucometric reporting in the latter half, adding further heterogeneity.

The concept of glucometrics is gaining more importance in inpatient settings, and reaching consensus is crucial for its wider application, as there are currently no standardized methods of monitoring diabetes care in inpatient settings, unlike well-established measurements such as HbA1c used for outpatient settings. In the United States, the Centers for Medicare & Medicaid Services, through the electronic clinical quality measures of hypoglycemia (<40 mg/dL in a patient-stay model) and hyperglycemia (>300 mg/dL in a patient-day model),⁵⁵ has been using glucometric data for institutional funding since 2021.¹⁴ The U.S. Centers for Disease Control and Prevention’s National Healthcare Safety Network are leading the efforts in developing digital quality measures for inpatient glucometrics, and are poised to generate standardized reportable metrics. In addition, further studies are required to identify which glucometric models best correlate with clinical outcomes such as mortality and health care-associated infections. Each glucometric model and performance metric has its unique advantages, and current glucometric reports include combination of models and performance metrics providing comprehensive understanding of specific glycemic trend. However, to maximize the value of glucometric analysis and benchmarking, consensus on reporting methods is essential, similar to advancements made for CGM devices in the ambulatory glucose profile.⁵⁶ Our review highlights sources of glucometric heterogeneity, affecting models, performance metrics, glucose thresholds, measurement methods, and exclusion criteria.

Glucometric Models and Performance Metrics

In our review, there was variability in the usage of glucometric models. While the majority (n=33) of studies preferred the “patient-day” model, highlighting preference for this more balanced model,¹³ many studies also used “patient-stay” and “population” models. While some heterogeneity is to be expected given the different uses, relying solely on less commonly used glucometric models can hinder comparison with other cohorts and studies. To enable direct comparisons between studies, we recommend that researchers report the more commonly used “patient-day” and “patient-stay” models as a bare minimum, in addition to other specific models as appropriate.

Further challenges arise from variability in performance metrics affecting how glucometric analyses are reported. Performance metrics used by Goldberg et al were “mean glucose” and “proportion above and below hyper- and hypoglycemia,” which were the most popular choices.¹³ Choice of performance metrics is further complicated by studies using novel metrics (i.e., adverse glycemic days, time-weighted average glucose, and glucose excursion rates) that provide valuable insights into minimizing hospital-acquired complications in patients with dysglycemia. However, these novel performance metrics present another source of heterogeneity limiting comparison across different cohorts and studies.

“Time-weighted average glucose” assesses the duration of time over which a patient experiences adverse glycemia, evaluating whether the length of each episode of adverse glycemia has greater prognostic value and a stronger association with mortality than existing metrics.⁵⁷ On the other hand, “adverse glycemic day” (AGD), defined as any patient-day with BG >270 mg/dL (>15 mmol/L) or <72 mg/dL (<4 mmol/L),³ indicates prevalence of the extreme ranges of BG to be avoided in hospital patients.³ AGDs can be considered the inverse of the “good diabetes day,” a term coined in the National Diabetes Inpatient Audit–England nationwide benchmarking study, and defined as a patient-day without hypoglycemia and with no more than one BG measurement exceeding 196 mg/dL (11 mmol/L).⁵⁸

Glucose Thresholds

A further barrier to glucometric harmonization is variation in glucose thresholds defining hyperglycemia and hypoglycemia. Our review shows that hypoglycemia thresholds range between 40 and 70 mg/dL (2.2-3.9 mmol/L) and hyperglycemia between 140 and 432 mg/dL (7.8-24.0 mmol/L). Some of this discrepancy may relate, in part, to different glucose targets in non-critical care and critical care settings, as well as differing targets between jurisdictions (Table 4).^57-61,64

Table 4.

Examples of jurisdictions with varying recommendations of inpatient glycemic targets.

Professional societies	Recommended glycemic targets
Professional societies	Non-critical Care	Critical Care
American Diabetes Association (ADA)	100–180 mg/dL (5.6–10.0 mmol/L)	Target range of 140-180 mg/dL (7.8-10.0 mmol/L)
American Association of Clinical Endocrinologists (AACE)	Fasting glucose < 140 mg/dL (< 7.8 mmol/L)Random glucose < 180 mg/dL (<10.0 mmol/L)	Target range of 140-180 mg/dL (7.8-10.0 mmol/L)
Endocrine Society	Target range of 100-180 mg/dL (5.6-10.0 mmol/L)
Joint British Diabetes Society	Target range of 72-216 mg/dL (4.0-12.0 mmol/L)
Academy of Royal Colleges	Target range of 108-216 mg/dL (6.0-12.0 mmol/L) during pre-, peri-, and post-operative inpatient diabetes
Society of Critical Care Medicine		140-200 mg/dL (7.8-11.1 mmol/L)
Australian Diabetes Society	90-180 mg/dL (5.0-10.0 mmol/L)	139-200 mg/dL (7.7-11.1 mmol/L)

Additionally, measurement units also contribute to this variability, with mg/dL often using multiples of 10 or 50, while mmol/L uses multiples of 0.5 and 1.0. The optimal glucose level for treating hyperglycemia has not been established through prospective randomized controlled trials in either non-critical or critical care settings, contributing further to variability. Recent studies indicate that while excessively tight glycemia in hospital between 80-110 mg/dL (4.4-6.1 mmol/L) may not be beneficial,^62,63 there is no consensus on which hyperglycemia level necessitates intervention.

The patient clinical context may further complicate glycemic treatment thresholds. For instance, the management of those experiencing new hyperglycemia may differ from those with long-standing diabetes and multiple comorbidities. For the former group, clinicians generally aim for stricter target glucose levels, while the presence of significant medical comorbidities and frailty may necessitate more liberal glucose targets.

Given the numerous factors influencing treatment decisions, it is crucial to standardize definitions of thresholds of adverse glycemia and graded thresholds for hypo- and hyperglycemia to enable direct comparisons between studies and to benchmark institutions. We recommend establishing minimum standard metrics when reporting glucometrics that is consistent with target ranges for BG set out by guidelines for inpatient diabetes care.^59,64

We propose reporting patient-day mean glucose levels exceeding 180 mg/dL (10 mmol/L), along with proportion of patient-days with any level 1 (54-69 mg/dL [3.0-3.8 mmol/L]) and level 2 hypoglycemia (<54 mg/dL [<3.0 mmol/L])²⁰ Moreover, < 40 mg/dL (<2.2 mmol/L) is also a worthwhile consideration as it represents “never event” values, but, given the low event frequency, may be better reported within a patient-stay model. These represent the most commonly used glucometric model and thresholds, respectively, that reflect a balance between strict and liberal glycemia.

Glucose Acquisition Method

A further source of variability in the reporting of glucometrics is the method used to obtain glucose measurements. Capillary POC BG measures were utilized in the majority of studies included in this systematic review, with some using plasma glucose or whole blood. Plasma and blood gas glucose samples are typically collected irrespective of diabetes status, while capillary POC BG measures, performed predominantly in those with diabetes, better capture symptomatic hyperglycemia and hypoglycemia.¹³ Therefore, in glucometric analyses using POC measures, the addition of plasma measures may lead to artificially lower reported glucose concentrations. Additionally, some studies indicate overall capillary POC BG measures to be typically 0.5-1.0 mmol/L lower than plasma glucose results.⁶⁵ Most studies reviewed did not address these potential biases resulting from combining POC with plasma glucose.

Even the use of POC capillary glucose poses challenges. The accuracy of POC results with blood samples is lower in the intensive care unit,⁴ and different POC instruments can produce dissimilar results, impacting data quality when dual devices are used.⁶⁶ Additionally, hospitals vary in their practices of BG testing frequencies and thresholds.⁴A promising glucose acquisition method to consider for future glucometry is measurement of interstitial fluid glucose concentration via CGM.^67,68 However, while CGM technology seems promising and consensus statements regarding harmonized reporting of CGM data exist,^11,69 it is not currently ready to be widely deployed in hospitals due to cost, staff training, inconsistent trial efficacy, and lack of regulatory approval for in-hospital use.²Additionally, it will be important for future studies to differentiate between interstitial fluid glucose from CGM and capillary BG from POC testing.⁷⁰

Exclusion Criteria

Finally, there is also a need for consensus on exclusion criteria for target populations. Typically, data from patients with diabetic ketoacidosis, hyperosmolar hyperglycemic state, critical care patients, pediatric patients, and pregnant individuals should be analyzed separately, due to skewed data that results from these patient populations having different glycemic measures in hospital or separate guidelines on target glycemic ranges.¹⁴ Additionally, data from palliative patients or patients with less than 24-hour hospitalizations should be excluded.¹⁴ Furthermore, including repeat glucose values after a hypoglycemic episode can skew results, as noted by Weinberg et al.³⁴ Therefore, it is essential to accurately report the specific exclusion criteria used. Despite the recommendations from Weinberg et al.³⁴, only three studies explicitly applied these exclusions.^3,5,9

Strengths and Limitations

The methodological strengths of this review include a comprehensive search strategy covering an extensive time period and a detailed review of multiple glucometric aspects. Although many studies were observational, given the descriptive nature of this review, this did not significantly impact our conclusions. An important limitation of our review is potential bias from including studies from institutions with greater access to glucometric data collection tools, including electronic medical records and connectivity-enabled glucose meters. Additionally, this study did not assess indices of glycemic variability, such as standard deviation, coefficient of variation, or mean amplitude of glycemic excursions. There is also growing interest in “insulinometrics,” which associates glucometrics with insulin use to assess inpatient glycemic management.^2,14 However, its accuracy and use rely on first achieving harmonization of glucometrics.

Conclusion

Harmonizing glucometrics enables direct performance comparisons within and between institutions, driving quality improvement in inpatient diabetes care. Key challenges include large data requirements, inconsistent glucose thresholds, and varied models and metrics, making comparisons difficult.

This review underscores the need to harmonize glucometric reporting to improve research comparability, enable benchmarking across hospitals, and support quality improvement initiatives.

Supplemental Material

sj-docx-1-dst-10.1177_19322968261427437 – Supplemental material for “Mellitus Metrics”—Systematic Review of Glucometric Reporting within Hospital-Based Diabetes Studies (2006-2023)

Supplemental material, sj-docx-1-dst-10.1177_19322968261427437 for “Mellitus Metrics”—Systematic Review of Glucometric Reporting within Hospital-Based Diabetes Studies (2006-2023) by Ji Won Susie Yoo, Ray Wang, Mervyn Kyi, Spiros Fourlanos and Rahul D Barmanray in Journal of Diabetes Science and Technology

Footnotes

Acknowledgements

none

Abbreviations

AGD, adverse glycemic day; BG, blood glucose; CGM, continuous glucose monitoring; HbA1c, hemoglobin A1c; POC, point-of-care; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Declaration of Conflicting Interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: RB chaired an educational meeting sponsored by an independent educational grant from Boehringer Ingelheim. MK has received honorarium from Astra Zeneca. SF contributes to the advisory panel for Viatris Inc. and Pfizer Inc; has contributed to the speaker’s bureau for Novo Nordisk, Astra Zeneca, and Boehringer Ingelheim and Eli Lilly Alliance; and has received research funding from Dexcom Inc. RW has received research funding from Dexcom Inc.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: RW is the recipient of an Australian Commonwealth Government RTP Scholarship, ACADI PhD Grant, Fred Knight Research Scholarship, Rowden White Scholarship, and Gordon P Castles Scholarship.

ORCID iDs

Ji Won Susie Yoo

Ray Wang

Rahul D Barmanray

Supplemental Material

Supplemental material for this article is available online.

References

Zhang

Bullard

Imperatore

Holliday

Benoit

SR.

Proportions and trends of adult hospitalizations with Diabetes, United States, 2000–2018. Diabetes Res Clin Pract. 2022;187: 109862. doi:10.1016/j.diabres.2022.109862.

Thompson

Cook

CB.

Glucometrics and insulinometrics. Curr Diab Rep. 2017;17(12):121. doi:10.1007/s11892-017-0964-2.

Kyi

Colman

Rowan

Marley

Wraight

Fourlanos

Glucometric benchmarking in an Australian hospital enabled by networked glucose meter technology. Med J Aust. 2019;211(4):175-180. doi:10.5694/mja2.50247.

Maynard

Schnipper

Messler

, et al. Design and implementation of a web-based reporting and benchmarking center for inpatient glucometrics. J Diabetes Sci Technol. 2014;8(4):630-640. doi:10.1177/1932296814532237.

Barmanray

Kyi

Worth

, et al. Hyperglycemia in hospital: an independent marker of infection, acute kidney injury, and stroke for hospital inpatients. J Clin Endocrinol Metabol. 2024;109(11):e2048-e2056. doi:10.1210/clinem/dgae051.

Dhatariya

Umpierrez

. Management of diabetes and hyperglycemia in hospitalized patients. Nih.gov. MDText.com, Inc.; 2024. https://www.ncbi.nlm.nih.gov/books/NBK279093/

Rubens

Ramamoorthy

Saxena

McGranaghan

McCormack-Granja

Recent trends in diabetes-associated hospitalizations in the United States. J Clin Med. 2022;11(22):6636. doi:10.3390/jcm11226636.

Barmanray

Kyi

Colman

Fourlanos

Longitudinal digital glucometric benchmarking to evaluate the impact of institutional diabetes care initiatives in adults with diabetes mellitus over the 2016-2020 period. J Diabetes Sci Technol. 2022;18(3):610-617. doi:10.1177/19322968221140126.

Kyi

Colman

Wraight

, et al. Early intervention for diabetes in medical and surgical inpatients decreases hyperglycemia and hospital-acquired infections: a cluster randomized trial. Diabetes Care. 2019;42(5):832-840.

10.

Saulnier

Castro

Cook

CB.

Use of cross-sectional and perspective mapping to spatially and statistically represent inpatient glucose control. J Diabetes Sci Technol. 2022;16(6):1385-1392.

11.

Thomas

P, E

Inzucchi

An internet service supporting quality assessment of inpatient glycemic control. J Diabetes Sci Technol. 2008;2(3):402-408. doi:10.1177/193229680800200309.

12.

Battelino

Danne

Bergenstal

, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range. Diabetes Care. 2019;42(8):1593-1603. https://doi.org/10.2337/dci19-0028.

13.

Goldberg

Bozzo

Thomas

, et al. “Glucometrics”–assessing the quality of inpatient glucose management. Diabetes Technol Ther. 2006;8(5):560-569. doi:10.1089/dia.2006.8.560.

14.

Engle

Bacani

Cook

Maynard

Messler

Kulasa

Glucometrics: where are we now. Curr Diab Rep. 2023;23(7):127-134. doi:10.1007/s11892-023-01507-1.

15.

Sly

Russell

Sullivan

Digital interventions to improve safety and quality of inpatient diabetes management: a systematic review. Int J Med Inform. 2022;157. doi:10.1016/j.ijmedinf.2021.104596.

16.

Barmanray

Kyi

Colman

, et al. The specialist treatment of inpatients: caring for diabetes in surgery (STOIC-D Surgery) trial: a randomized controlled trial of early intervention with an electronic specialist-led model of diabetes care. Diabetes Care. 2024;2024:dc231905. doi:10.2337/dc23-1905.

17.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi:10.1136/bmj.n71.

18.

Maynard

Holdych

Kendall

Harrison

Montgomery

Kulasa

Improving glycemic control safely in critical care patients: a collaborative systems approach in nine hospitals. Endocr Pract. 2017;23(5):583-593. doi:10.4158/EP161569.OR.

19.

Maynard

Childers

Holdych

Kendall

Hoag

Harrison

Improving glycemic control safely in non-critical care patients: a collaborative systems approach in nine hospitals. Jt Comm J Qual Patient Saf. 2017;43(4):179-188. doi:10.1016/j.jcjq.2017.01.003.

20.

McIver

Mitchell

Finn

Kamp

MC.

Standardising practices through form design and education improves insulin management. Aust Health Rev. 2009;33(3):434-441. doi:10.1071/ah090434.

21.

Cheung

Cinnadaio

O’Neill

, et al. Implementation of a dedicated hospital subcutaneous insulin prescription chart: effect on glycaemic control. Diabetes Res Clin Pract. 2011;92(3):337-341. https://doi.org/10.1016/j.diabres.2011.02.019.

22.

Bogun

Beier

Singh

, et al. Diabetes workshops for providers improve glucose control in coronary artery bypass grafting patients. J Card Surg. 2022;37(4):930-936. doi:10.1111/jocs.16282.

23.

Lleva

Thomas

Bozzo

Hendrickson

Inzucchi

SE.

Using the glucometrics website to benchmark ICU glucose control before and after the NICE-SUGAR study. J Diabetes Sci Technol. 2014;8(5):918-922. doi:10.1177/1932296814540871.

24.

Shelden

Ateya

Jensen

Arnold

Bellomo

Gianchandani

Improving hospital glucometrics, workflow, and outcomes with a computerized intravenous insulin dose calculator built into the electronic health record. J Diabetes Sci Technol. 2021;15(2):271-278. doi:10.1177/1932296820974767.

25.

Shabbir

Stein

Tong

Bhalla

Wang

A real-time nursing intervention reduces dysglycemia and improves best practices in noncritically ill hospitalized patients. J Hosp Med. 2010;5(1):E15-20. doi:10.1002/jhm.590.

26.

Bansal

Mithal

Carvalho

Mehta

Trehan

Feasibility, efficacy, and safety of a simple insulin infusion protocol in a large volume cardiac surgery unit in India. Indian J Endocrinol Metab. 2015;19(1):47-51. doi:10.4103/2230-8210.146864.

27.

Maynard

Lee

Phillips

Fink

Renvall

Improved inpatient use of basal insulin, reduced hypoglycemia, and improved glycemic control: effect of structured subcutaneous insulin orders and an insulin management algorithm. J Hosp Med. 2009;4(1):3-15. doi:10.1002/jhm.391.

28.

Kao

Chen

Ning

, et al. Evaluating the effectiveness of a multi-faceted inpatient diabetes management program among hospitalised patients with diabetes mellitus. Clin Diabetes Endocrinol. 2020;6(1):21. doi:10.1186/s40842-020-00107-2.

29.

Brand

Peragallo-Dittko

Fazzari

Islam

Jacobson

Radin

MS.

Changing to basal-bolus insulin therapy for the inpatient management of hyperglycemia-a natural experiment. Endocr Pract. 2019;25(8):836-845. doi:10.4158/EP-2018-0498.

30.

Mendez

Ata

Rourke

Stain

Umpierrez

Daily inpatient glycemic survey (dings): a process to remotely identify and assist in the management of hospitalized patients with diabetes and hyperglycemia. Endocr Pract. 2015;21(8):927-935. doi:10.4158/EP14577.OR.

31.

Engle

Ferguson

Fields

A journey to improved inpatient glycemic control by redesigning meal delivery and insulin administration. Clin Nurse Spec. 2016;30(2):117-124. doi:10.1097/NUR.0000000000000190.

32.

Goswami

Scheinberg

Schechter

Ruocco

Davis

NJ.

Impact of multidisciplinary process improvement interventions on glucometrics in a noncritically ill setting. Endocr Pract. 2019;25(7):689-697. doi:10.4158/EP-2018-0497.

33.

Mulla

Lieb

McFarland

Aloi

JA.

Tides of change: improving glucometrics in a large multihospital health care system. J Diabetes Sci Technol. 2015;9(3):602-608. doi:10.1177/1932296814563953.

34.

Weinberg

Bacchetti

Rushakoff

RJ.

Frequently repeated glucose measurements overestimate the incidence of inpatient hypoglycemia and severe hyperglycemia. J Diabetes Sci Technol. 2010;4(3):577-582. doi:10.1177/193229681000400311.

35.

Koziol

Johnson

Brenner

Fortmann

Morrisey

Philis-Tsimikas

Novel approach to inpatient glucometric monitoring and variability in a community hospital setting. J Diabetes Sci Technol. 2015;9(2):246-256. doi:10.1177/1932296814564992.

36.

Kosiborod

Inzucchi

Krumholz

, et al. Glucometrics in patients hospitalized with acute myocardial infarction: defining the optimal outcomes-based measure of risk. Circulation. 2008;117(8):1018-1027. doi:10.1161/CIRCULATIONAHA.107.740498.

37.

Buchs

Rapoport

MJ.

Institutional glucometrics to determine glucose control as practiced by general medicine wards. Isr Med Assoc J. 2010;12(8):463-467.

38.

Dharma

JFM

Polmear

Fourlanos

Kyi

. Reporting hypoglycaemia as a hospital-acquired complication: assessing biochemical and clinical validity. Intern Med J. 2024;54(3):430-437. doi:10.1111/imj.16124. (Published online ahead of print on 15 May 2023).

39.

Guerra

Das

Antonopoulos

, et al. Computerized physician order entry-based hyperglycemia inpatient protocol and glycemic outcomes: the CPOE-HIP study. Endocr Pract. 2010;16(3):389-397. doi:10.4158/EP09223.OR.

40.

Khatib

Borawake

Glucometrics of diabetic patients admitted to intensive care unit in hospitals with limited information technology support: is it possible. J Diabetes Sci Technol. 2014;8(5):1055-1056. doi:10.1177/1932296814535732.

41.

Bersoux

Cook

Kongable

Shu

Zito

DR.

Benchmarking glycemic control in U.S. Endocr Pract. 2014;20(9):876-883. doi:10.4158/EP13516.OR.

42.

Saulnier

Castro

Cook

Thompson

BM.

How is the weather? Forecasting inpatient glycemic control. Future Sci OA. 2017;3(4):FSO241. doi:10.4155/fsoa-2017-0066.

43.

Bersoux

Cook

Kongable

Shu

Trends in glycemic control over a 2-year period in 126 US hospitals. J Hosp Med. 2013;8(3):121-125. doi:10.1002/jhm.1997.

44.

Rao

Perreiah

Cunningham

CA.

Monitoring the impact of aggressive glycemic intervention during critical care after cardiac surgery with a glycemic expert system for nurse-implemented euglycemia: the MAGIC GENIE project. J Diabetes Sci Technol. 2021;15(2):251-264. doi:10.1177/1932296821995568.

45.

Munshi

Saghafian

Cook

Werner

Chakkera

HA.

Comparison of post-transplantation diabetes mellitus incidence and risk factors between kidney and liver transplantation patients. PLoS ONE. 2020;15(1):e0226873. doi:10.1371/journal.pone.0226873.

46.

Schnipper

Barsky

Shaykevich

Fitzmaurice

Pendergrass

ML.

Inpatient management of diabetes and hyperglycemia among general medicine patients at a large teaching hospital. J Hosp Med. 2006;1(3):145-150. doi:10.1002/jhm.96.

47.

Munoz

Pronovost

Dintzis

, et al. Implementing and evaluating a multicomponent inpatient diabetes management program: putting research into practice. Jt Comm J Qual Patient Saf. 2012;38(5):195-206. doi:10.1016/s1553-7250(12)38025-2.

48.

Munshi

Saghafian

Cook

Steidley

Hardaway

Chakkera

HA.

Incidence, risk factors, and trends for postheart transplantation diabetes mellitus. Am J Cardiol. 2020;125(3):436-440. doi:10.1016/j.amjcard.2019.10.054.

49.

Seheult

Pazderska

Gaffney

, et al. Addressing inpatient glycaemic control with an inpatient glucometry alert system. Int J Endocrinol. 2015;2015:807310. doi:10.1155/2015/807310.

50.

Chen

Ning

Thomas

, et al. An open source tool to compute measures of inpatient glycemic control: translating from healthcare analytics research to clinical quality improvement. JAMIA Open. 2021;4(2):ooab033. doi:10.1093/jamiaopen/ooab033.

51.

Rushakoff

Sullivan

Seley

, et al. Using a mentoring approach to implement an inpatient glycemic control program in United States hospitals. Healthc (Amst). 2014;2(3):205-210. doi:10.1016/j.hjdsi.2014.06.004.

52.

Schnipper

Liang

Ndumele

Pendergrass

ML.

Effects of a computerized order set on the inpatient management of hyperglycemia: a cluster-randomized controlled trial. Endocr Pract. 2010;16(2):209-218. doi:10.4158/EP09262.OR.

53.

Bernard

Munoz

Harper

Muriello

Rico

Baldwin

Treatment of inpatient hyperglycemia beginning in the emergency department: a randomized trial using insulins aspart and detemir compared with usual care. J Hosp Med. 2011;6(5):279-284. doi:10.1002/jhm.866.

54.

Lin

Jhang

Hsieh

MC.

A workable model for the management of hyperglycemia in non-critically ill patients in an Asian population. Postgrad Med. 2015;127(8):796-800. doi:10.1080/00325481.2015.1080113.

55.

eCQI Resource Center. Hospital—Inpatient eCQMs; 2024. https://ecqi.healthit.gov/eh-cah/ecqms. Accessed December 7, 2025.

56.

Battelino

Danne

Bergenstal

57.

American Diabetes Association Professional Practice Committee; 16. Diabetes care in the hospital: standards of care in diabetes—2025. Diabetes Care. 2025;48 (Supplement_1):S321–S334. Doi:10.2337/dc25-s016.

58.

Digital NHS. National Diabetes Inpatient Audit (NaDIA) 2016. March 2017. http://www.digital.nhs.uk/pubs/nadia2016. Accessed May 31, 2025.

59.

Joint British Diabetes Societies for inpatient care. A good inpatient diabetes service (Guideline No. 14), 2019. https://abcd.care/sites/default/files/site_uploads/JBDS_Guidelines_Current/JBDS_14_A_Good_Inpatient_Service_Updated_060720.pdf.

60.

Academy of Medical Royal Colleges Centre for Perioperative Care. Guideline for Perioperative Care for People with Diabetes Mellitus Undergoing Elective and Emergency Surgery. London: Centre for Perioperative Care; 2023.

61.

Korytkowski

Muniyappa

Antinori-Lent

, et al. Management of hyperglycemia in hospitalized adult patients in non-critical care settings: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2022;107(8):2101-2128. doi:10.1210/clinem/dgac278.

62.

Van den Berghe

Schetz

Vlasselaers

, et al. Intensive insulin therapy in critically ill patients: NICE-SUGAR or Leuven blood glucose target? J Clin Endocrinol Metabol. 2009;94(9): 3163170. doi:10.1210/jc.2009-0663.

63.

Gunst

Debaveye

Güiza

, et al. Tight blood-glucose control without early parenteral nutrition in the ICU. New Engl J Med. 2023;389(13):1180-1190. DOI:10.1056/NEJMoa2304855.

64.

Australian Diabetes Society. Guidelines for Routine Glucose Control in Hospital; 2012. https://www.diabetessociety.com.au/documents/ADSGuidelinesforRoutineGlucoseControlinHospitalFinal2012.pdf.

65.

Carstensen

Lindström

Sundvall

Borch-Johnsen

Tuomilehto

; DPS Study Group. Measurement of blood glucose: comparison between different types of specimens. Ann Clin Biochem. 2008;45(Pt2):140-148. doi:10.1258/acb.2007.006212.

66.

Ekhlaspour

Mondesir

Lautsch

, et al. Comparative accuracy of 17 point-of-care glucose meters. J Diabetes Sci Technol. 2017;11(3):558-566. doi:10.1177/1932296816672237.

67.

Wang

Krishnamoorthi

Connell

, et al. Rapid increase in continuous glucose monitoring use among adults with type 1 diabetes admitted to hospital: a Melbourne multicentre observational study. Intern Med J. 2025;55(7):1184-1187. doi:10.1111/imj.70095.

68.

Wang

Kyi

Krishnamoorthi

, et al. Accuracy of continuous glucose monitoring in people with type 1 diabetes receiving hemodialysis in hospital. J Diabetes Sci Technol. 2025;19(3):859-861. doi:10.1177/19322968251318758.

69.

Battelino

Alexander

Amiel

, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol. 2023;11(1):42-57. https://doi.org/10.1016/S2213-8587(22)00319-9.

70.

Espinoza

Klonoff

Vidmar

, et al. 2022 iCoDE Report: CGM-EHR integration standards and recommendations. Diabetes Technology Society; November 7, 2022. https://www.diabetestechnology.org/icode/. Accessed June 6, 2025.