Dear Editor,

The accuracy of a point-of-care (POC) glucometer is closely related to the quality of insulin dosing decisions [1, 2]. The performance of a POC glucometer should be sufficient for a pre-market study. However, marketed devices do not always meet requirements [3, 4]. Pre-market studies have limitations in detecting uncommon adverse events (AEs), long-term complications, AEs associated with off-label use, and post-market issues [5]. Therefore, the need for obtaining real-world evidence (RWE) such as AE reporting derived from multiple sources outside typical clinical research settings has been emphasized. The European Medicines Agency suggested three options for monitoring AEs [6]. Among them, active surveillance and electronic health record (EHR) analysis have been applied in previous studies of POC glucometer performance [3, 7]. However, enhanced passive surveillance (EPS) has not been reported.

We aimed to promote the quantity and quality of AE reporting related to the use of POC glucometers in a single center and to review related data from the Ministry of Food and Drug Safety (MFDS) of Korea and the United States Food and Drug Administration (USFDA).

Barozen H Expert Plus (i-SENS, Seoul, Korea) was used as the POC glucometer at Soonchunhyang University Bucheon Hospital (SCHBC), Bucheon, Korea. Combined with routine surveillance, the EPS mandate of AE reporting (Fig. 1) has been adopted since October 1, 2021. The AE reports for 1 year were analyzed [8]. Databases of the MFDS’s Medical Device Information Portal (https://udiportal.mfds.go.kr/) and USFDA’s Total Product Life Cycle (https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfTPLC/tplc.cfm) were used to search for AE reports from 2017 to September 2022. The product codes for the POC glucometers used in Korea and the United States were obtained from the blood glucose test system and NBW (System, Test, Blood Glucose, Over The Counter), respectively. Collected AEs were classified using International Medical Device Regulators Forum (IMDRF) codes for medical device problems (Annex A) and clinical signs and symptoms or conditions (Annex E) [9, 10]. This study did not require Institutional Review Board approval, because it did not involve human subjects, including any interaction or intervention with human subjects, or access to identifiable private information.

Figure 1. Enhanced passive surveillance applied to POC glucometer adverse event reporting; (A) Overview and (B) IMDRF code conversion of reported adverse events using checklist.
Abbreviations: POC, point-of-care; QC, quality control; IMDRF, International Medical Device Regulator Forum.

Only three AEs of the POC glucometer were reported to the Medical Device Safety Monitoring Center of SCHBC (unpublished data) from 2013 to 2021. However, EPS enabled identifying 137 AE reports with eight Annex A codes and one Annex E code for 1 year. Five Annex A codes (A070908, failure to sense; A1302, failure to transmit record; A1303, intermittent communication failure; A150105, key or button unresponsive/not working; and A180103, device contamination with body fluid), which accounted for 70.9% (97/137) of the AE reports were reported in only the SCHBC database and not in the two AE databases.

Unlike SCHBC, the AE database showed various codes (97 Annex A codes and 106 Annex E codes) and a large number of reports (12 AEs in the MFDS database and 94,126 AEs in the USFDA database) (Table 1). Codes under A0908 (incorrect, inadequate, or imprecise result or readings), including A090805 (incorrect measurement) and A090809 (high test results), have been frequently reported by operators after finding erroneous results of POC glucometers in Korea and the United States. Inaccurate test results leading to inadequate treatment would result in codes E1205 (hyperglycemia) and E1206 (hypoglycemia) [2].

Table 1 . POC glucometer-related adverse events reported by SCHBC from October 2021 to September 2022 and registered in the MFDS and USFDA databases since 2017

IMDRF Code			Definition	SCHBC (N=137), N (%)	MFDS (N=12), N (%)	USFDA (N=94,126), N (%)
Lv 1	Lv 2	Lv 3				Cumulated	Reported
Annex A*: Medical device problem
A02			Manufacturing, packaging, or shipping problem	NR	NR	1,157 (1.2)	14 (0.0)
	A0201		Product quality problem	NR	NR	100 (0.1)	53 (0.1)
		A020102	Nonstandard device	NR	1 (8.3)	47 (0.0)	47 (0.0)
A07			Electrical/electronic property problem	NR	NR	40,211 (41.2)	360 (0.4)
	A0705		Battery problem	NR	NR	1,546 (1.6)	1,524 (1.6)
	A0708		Power problem	NR	NR	38,256 (39.2)	28 (0.0)
		A070803	Failure to power up	10 (7.3)	NR	38,217 (39.1)	38,217 (39.1)
	A0709		Device sensing problem	NR	2 (16.7)	30 (0.0)	30 (0.0)
		A070908	Failure to sense	70 (51.1)	NR	NR	NR
A09			Output problem	NR	NR	45,615 (46.7)	567 (0.6)
	A0902		Display or visual feedback problem	5 (3.6)	NR	13,944 (14.3)	172 (0.2)
		A090201	Device displays incorrect message	NR	2 (16.7)	10,973 (11.2)	1,0973 (11.2)
		A090202	Display difficult to read	NR	NR	1,388 (1.4)	1,388 (1.4)
	A0908		Incorrect, inadequate, or imprecise result or readings	18 (13.1)	NR	30,643 (31.4)	22,253 (22.8)
		A090805	Incorrect measurement	NR	5 (41.7)	56 (0.1)	56 (0.1)
		A090807	High readings (A090809 is more recommended)	NR	1 (8.3)	1,336 (1.4)	1,336 (1.4)
		A090808	Low readings (A090810 is more recommended)	NR	1 (8.3)	438 (0.4)	438 (0.4)
		A090809	High test results	NR	NR	3,915 (4.0)	3,915 (4.0)
		A090810	Low test results	NR	NR	1,406 (1.4)	1,406 (1.4)
A13			Communication or transmission problem	NR	NR	26 (0.0)	12 (0.0)
	A1302		Failure to transmit record	12 (8.7)	NR	NR	NR
	A1303		Intermittent communication failure	4 (3.0)	NR	NR	NR
A15			Activation, positioning, or separation problem	NR	NR	NR	NR
	A1501		Activation problem	NR	NR	NR	NR
		A150105	Key or button unresponsive/not working	1 (0.7)	NR	NR	NR
A16			Protective measures problem	NR	NR	1,649 (1.7)	7 (0.0)
	A1601		Device alarm system	NR	NR	1,642 (1.7)	1,642 (1.7)
A18			Contamination/decontamination problem	NR	NR	NR	NR
	A1801		Contamination	NR	NR	NR	NR
		A180103	Device contamination with body fluid	10 (7.3)	NR	NR	NR
A19			Environmental compatibility problem	NR	NR	2,489 (2.5)	NR
	A1910		Electrical power problem	NR	NR	2,489 (2.5)	NR
		A191002	Loss of power (A070803 is more recommended)	NR	NR	2,489 (2.5)	2,489 (2.5)
A23			Use of device problem	NR	NR	2,206 (2.3)	1,702 (1.7)
A27			Appropriate term/code not available	NR	NR	1,113 (1.1)	1,113 (1.1)
Annex E: Clinical signs and symptoms or conditions
E01			Nervous system	NR	NR	6,931 (7.4)	NR
	E0112		Dizziness	NR	NR	1,783 (1.9)	1,783 (1.9)
	E0119		Loss of consciousness	NR	NR	1,472 (1.6)	801 (0.9)
	E0122		Movement disorder	NR	NR	1,886 (2.0)	NR
		E012206	Shaking/tremors	NR	NR	1,841 (2.0)	1,841 (2)
E12			Endocrine, metabolism, and nutrition	NR	NR	4,667 (5.0)	NR
	E1205		Hyperglycemia	NR	1 (8.3)	1,826 (1.9)	1,675 (1.8)
	E1206		Hypoglycemia	NR	1 (8.3)	2,299 (2.4)	2,299 (2.4)
E23			Generalized disorders	NR	NR	2,779 (2.7)	NR
	E2310		Diaphoresis	NR	NR	1,653 (1.6)	1,653 (1.6)
E24			Others	NR	NR	84,567 (82.4)	NR
	E2403		No clinical signs, symptoms, or conditions	137 (100.0)	9 (75.0)	83,664 (81.5)	83,664 (81.5)

*Annex A codes, which have not been reported in SCHBC and Korea and have been reported to account for <1% of codes in the United States, were included in the total number of cases, but individual descriptions were excluded.

Abbreviations: POC, point-of-care; USFDA, United States Food and Drug Administration; MFDS, Ministry of Food and Drug Safety; SCHBC, Soonchunhyang University Bucheon Hospital; IMDRF, International Medical Device Regulators Forum; Lv, level; NR, not reported.

Many factors, such as old technology that may not meet current device standards, an inadequate pre-market study, and manufacturing issues, might have contributed to the inadequate performance. In addition to the inaccuracy of the POC glucometer itself, analytical factors, such as extreme physical environments, polycythemia, anemia, and certain medications can also adversely affect the accuracy of blood glucose monitoring [2]. To address this, constant patient monitoring and surveillance protocols are required. Although codes A070908 and A1302 in SCHBC were not observed in the AE database, they have been reported as potential sources of error in blood glucose monitoring [2]. Barozen Expert plus requires operator and patient identification barcode reading process for automatic transmission of test results to the EHR. We judged all barcode reading failures for operator, patients, and test strip identification to fall under code A070908, whereas code A1302 indicates failure to transmit records to the EHR.

This study had several limitations. First, EPS could mandate AE reporting when only the POC glucometer was replaced. Second, the success of EPS still depends on stakeholders’ interests’ and commitment to the AE report. Third, only the AE codes were evaluated. In particular, no health effects were reported in 1 year.

There were only three AEs reported related to the use of a POC glucometer by SCHBC in 9 years, but EPS has been effective in promoting AE reporting, increasing the number of reports to 137 during the 1-year study period after EPS application. Increased AE reporting through EPS can contribute to improving the safety of in vitro tests of medical devices and generating RWE.

The authors wish to acknowledge the contributions of Jae Jun Lee, M.T. and Yoon Jung Jin, R.N. of Soonchunhyang University Bucheon Hospital and Hyeran Oh and Youna Lee of Soonchunhyang Medical Device Clinical Research Center.

Choi S analyzed the data and wrote the manuscript; Choi SJ edited and revised the manuscript; and Choi SJ, Kim JK, Lee Y-W, and Lee YK collected the data. All authors reviewed and approved the final version of the manuscript.

Choi S conceived the study. Choi S and Choi SJ contributed to data interpretation. All authors contributed to manuscript writing

The authors declare no conflicts of interest.

This research was funded by the Soonchunhyang University Research Fund and the National Institute of Medical Device Safety Information (NIDS).