Acute Hepatitis B Outbreaks Related to Fingerstick Blood Glucose Monitoring in Two Assisted Living Facilities

Abstract

OBJECTIVES: To establish the etiology for outbreaks of hepatitis B virus (HBV) infections at two assisted living facilities (ALFs) and devise appropriate control measures.

DESIGN: Multisite outbreak investigations, retrospective cohort.

SETTING: Two ALFs in Illinois.

PARTICIPANTS: Facility A residents (n=120) and Facility B residents (n=105) and nursing staff (n=6).

MEASUREMENTS: For Facility A, a retrospective cohort study to identify risk factors for HBV infection through serological testing of all residents and a medical record extraction. For Facility A and B, investigation of fingerstick blood glucose monitoring techniques. For Facility B, serological HBV testing of nurses and residents receiving fingerstick blood glucose monitoring.

RESULTS: At Facility A, five confirmed acute, two probable acute, and one probable chronic HBV infections were identified in the 109 residents tested. All of the eight identified residents with HBV infection had diabetes mellitus. HBV deoxyribonucleic acid (DNA) sequences from the chronic and acute cases were identical. Transmission of HBV was associated with fingerstick blood glucose monitoring (relative risk (RR)=28.5, 95% confidence interval (CI)=1.6–498; P<.001) and insulin injections (RR=7.4, 95% CI=1.3–40.8; P=.03). At Facility B, seven of 21 residents (33.3%) receiving fingerstick blood glucose monitoring had evidence of recent HBV infection.

CONCLUSION: Nurses probably transmitted HBV infection from resident to resident during fingerstick blood glucose monitoring in two separate ALFs, causing outbreaks. Awareness of the high risk for HBV transmission during procedures for the care of diabetes mellitus was limited. Following established infection control measures is critical to prevent spread of this highly contagious virus.

Outbreaks of hepatitis B virus (HBV) infection have occurred in residents of long-term care facilities (LTCFs) with diabetes mellitus as a result of inappropriate sharing of equipment and inadequate aseptic technique during fingerstick blood glucose monitoring.^1–3 Accumulating evidence suggests that such transmission may be more widespread than previously recognized, particularly in the assisted living environment.⁴

HBV is highly infectious and easily transmitted after percutaneous exposure. In studies of healthcare professionals who sustained injuries from needles contaminated with blood containing HBV, the estimated risk of developing serological evidence of HBV infection was 23% to 62%, far greater than that from exposure to hepatitis C virus (1.8%) or human immunodeficiency virus (0.3%).⁵ Similarly, the potential for transmission of HBV during blood glucose monitoring has long been established.⁶ HBV concentrations in blood can be high, and HBV is known to remain viable on contaminated surfaces for 1 week or longer.⁷ Unless meticulous attention is paid to preventing inadvertent environmental contamination with blood, fingerstick blood glucose monitoring can place diabetic residents at substantial risk for exposure to HBV and other bloodborne pathogens.

Depending on state- and facility-specific requirements, assisted living facilities (ALFs) can accommodate resident-performed and staff-assisted blood glucose monitoring. Typically, there is less regulation of infection control measures and assurance of trained personnel in ALFs than in facilities licensed as nursing homes.^8–10

Delayed recognition of HBV transmission or missed acute HBV infections within a facility presents challenges to identifying the source of infection and prevention of additional cases. Elderly adults have few typical risk factors for infection (e.g., illicit drug use or high-risk sexual behaviors), and healthcare providers are often not familiar with the potential for HBV transmission during routine care for diabetes mellitus.^11,12

METHODS

In each outbreak, an epidemiological investigation was undertaken to determine the likely mode of transmission and recommend appropriate preventive measures.

Inspection

On March 7, 2007, a hospital infection control professional (ICP) notified the Cook County Department of Public Health (CCDPH) of two patients with diabetes mellitus hospitalized with acute HBV infection. Both were residents of a 129-bed ALF (Facility A) in suburban Cook County housing 120 residents and employing three nurses (2 licensed practical nurses (LPNs) and 1 registered nurse (RN), who served as the Director of Nursing (DON)). CCDPH investigators visited the facility to review infection control procedures with the DON and other staff.

On April 8, 2008, CCDPH was notified of a patient with diabetes mellitus hospitalized with acute HBV infection. The patient resided at a 120-bed four-story ALF (Facility B) in suburban Cook County housing 105 residents and employing seven nurses (1 part-time RN consultant, 4 full-time and 2 part-time LPNs). This was the second recognized case of acute HBV infection in a Facility B resident in a 15-month period. For the second time, CCDPH investigators visited Facility B to review infection control procedures with the DON and other staff.

Walk-through on-site observations were performed in both facilities, with an emphasis on identifying potential sources of bloodborne pathogen exposure.

Retrospective Cohort Study

A retrospective cohort study was conducted at Facility A to identify cases and risk factors for HBV infection. The exposure period for this outbreak was estimated to be between June 1, 2006, and December 31, 2006. Of the 120 residents, 11 initially tested for HBV entered the facility after the probable exposure period and were excluded from the study. Data were abstracted from medical records using a standardized data collection instrument. Variables collected from residents included demographic information, date admitted to the facility, history of diabetes mellitus, frequency of fingerstick blood glucose monitoring, insulin injection, laboratory test results, and prior history of HBV infection. Information about procedures with the potential for bloodborne pathogen exposure, including dates, was tabulated. These potential exposures included receiving on-site services by contract providers (podiatrist, dentist, and laboratory phlebotomist) and hospitalizations.

Serologic Testing

All 120 residents of Facility A at the time of the investigation were offered serological testing; facility nurses were not tested. At Facility B, it was not possible to establish an exposure period. The 21 residents who were present and who had received fingerstick blood glucose monitoring any time after January 1, 2006, had serological testing for HBV. All six facility nurses who performed fingerstick blood glucose monitoring were tested. A local community laboratory performed serological testing for the presence of hepatitis B surface antigen (HBsAg). Positive specimens were then tested for total and immunoglobulin (IgM) antibody to hepatitis B core antigen (anti-HBc) and antibody to HBsAg (anti-HBs). Residents of Facility A who were HBsAg positive were offered repeat testing 6 months after the initial investigation to help assess whether the infection had been cleared or become chronic.

Sera from HBsAg-positive residents from Facility A were submitted to the Centers for Disease Control and Prevention (CDC) Viral Hepatitis Molecular Epidemiology Laboratory for HBV deoxyribonucleic acid (DNA) sequencing and phylogenetic analysis.

Statistical Analysis

Relative risks (RRs), 95% confidence intervals (CIs), and P-values were calculated for the exposures at Facility A using SAS v9.13 (SAS Institute, Inc., Cary, NC). For variables with 0 cells, 0.5 was added to each cell to estimate the RR. If the expected cell count was less than 5, Fisher exact P-values were calculated (Table 1).

Table 1.   Risk of Acquiring Hepatitis B Virus (HBV) Infection in a Cohort of Residents of Assisted Living Facility A Between June 1 and December 31, 2006 (n=95)

Exposure	Confirmed Acute HBV Cases, n (%) n=5	Noncases, n (%) n=90	Relative Risk (95% Confidence Interval)	P-Value^*
^* Calculated using Fisher exact test on original cell counts. ^† ^† 0.5 added to each cell.
Diabetic^†	5 (100)	20 (22)	30.0 (1.7–525)	<.001
Fingerstick glucose^†	5 (100)	19 (21)	28.5 (1.6–498)	<.001
Insulin	3 (60)	13 (14)	7.4 (1.2–41)	.03
Podiatric care	4 (80)	46 (50)	3.6 (0.42–31)	.37
Dental care	3 (60)	7 (8)	12.8 (2.4–67)	.008
Hospitalization	2 (40)	26 (29)	1.6 (0.28–9.0)	.63
Blood draw	1 (20)	30 (33)	0.52 (0.06–4.4)	1.0

Record Review

At Facility B, select data were abstracted from the medical records of residents with diabetes mellitus. The investigation focused on fingerstick blood glucose monitoring, insulin injection, and specific nursing assignments.

CCDPH matched resident names with reported hepatitis B cases in the state of Illinois' National Electronic Disease Surveillance System to determine whether there had been other cases at the facility.

RESULTS

Facility A: Case Identification

At Facility A, 109 of the 120 residents were tested for serological markers of HBV infection. Eleven residents were not tested, because they refused (n=6), had died (n=1), were hospitalized at the time of the study (n=2), or had been discharged (n=2). The residents who died or were hospitalized did not have symptoms consistent with HBV infection. Thirty of the 32 residents (93.8%) receiving routine fingerstick blood glucose monitoring were tested; two of these residents did not have diabetes mellitus. No residents were known to have a prior history of HBV infection.

Eight of the 109 tested residents were HBsAg positive; all positive residents (100%) received fingerstick blood glucose monitoring. The community laboratory initially handling the specimens could not test for the presence of anti-HBc (IgM or total) to determine whether the infections were acute or chronic and discarded the specimens. Residents needed to have blood redrawn to test for IgM; two of the eight refused (probable acute cases). Repeat testing was immediately performed. Of the six remaining HBsAg-positive residents who were tested for anti-HBc IgM, five were positive (confirmed acute cases), and one was negative (probable chronic case).

In all, seven acute hepatitis B cases (5 confirmed, 2 probable) were identified at Facility A with symptom onset or diagnosis between January 5, 2007, and April 20, 2007. Three of the cases were hospitalized during this period. Of the remaining four cases, none were known to have symptoms consistent with HBV infection.

The ALF opened in August 2003, and the probable chronic hepatitis B case moved in at that time. The residents with acute HBV infection had lived at the facility for 1.6 to 4.6 years before diagnosis. Six cases had daily routine fingerstick blood glucose monitoring; the seventh case had this procedure performed three times per week.

One hundred one residents were HBsAg negative. Eleven of these residents were excluded from the cohort study after medical record extraction indicated that they had moved to the facility after the estimated exposure period.

Facility A: Retrospective Cohort Study

Ninety-five residents were included in the retrospective cohort study for Facility A, which compared the five residents with confirmed acute HBV infection with the remaining 90 residents who tested negative and had remained susceptible to HBV infection throughout the estimated exposure period. The two residents with probable acute HBV infection were excluded from the study; it could not be determined with certainty that their infections were acute. The results of the medical record extraction and statistical analysis are included in Table 1.

Residents with diabetes mellitus (RR=30.0, 95% CI=1.7–525, P<.001) or fingerstick blood glucose monitoring (RR=28.5, 95% CI=1.6–498, P<.001) were at the greatest risk of infection; all five confirmed cases, as well as the two probable cases, shared these characteristics. Insulin injection was also significantly associated with acute HBV infection (RR=7.4, 95% CI=1.2–41, P=.03). Three of the five confirmed cases received insulin. Insulin was administered at the same time as fingerstick testing was performed, and it was not possible to separate those two variables during the analysis. Including the two probable acute cases in the analysis further strengthened these associations (diabetes mellitus status: RR=42.2, 95% CI=2.2–644, P<.001; fingerstick blood glucose monitoring: RR=36.2, 95% CI=2.1–613, P<.001; and insulin injection: RR=11, 95% CI=2.3–52, P=.002).

Three case patients received dental care, and this exposure was statistically significant (RR=12.8, 95% CI=2.4–67; P=.008), although these results appeared to be an artifact related to the small number of cases. Of the eight HBsAg-positive residents at the facility, only three received dental care on site, and none received care on the same date. Moreover, the putative source patient with probable chronic HBV infection did not receive dental care at the facility.

Facility A: HBV DNA Sequence Analysis

Specimens from six HBsAg-positive residents (1 probable acute case, 4 confirmed acute cases, and 1 probable chronically infected resident) were submitted to the CDC for HBV DNA sequence analysis. The HBV genomes in the region (2,150 nucleotides) comprising the complete surface-antigen-coding gene and the X gene, as well as a significant part of polymerase gene, were sequenced. All six sequences were identical and belonged to HBV subgenotype A2, predicting the adw2 subtype (Figure 1). Two HBsAg-positive residents refused an additional blood draw for this test.

Figure 1.  Facility A hepatitis B virus (HBV) deoxyribonucleic acid (DNA) sequence analysis. All HBV genomic sequences of the five case patients' isolates were identical, supporting the hypothesis of transmission within the facility. Sequences in the 2,150 base pair PreS-S-X-PreCore genes region in samples from Facility A case patients were compared with reference sequences in the GenBank database. Each of the five case patients on whom HBV subtyping was performed had subgenotype A2, predicting the adw2 subtype, including the probable HBV carrier and the index patient. All sequences of the five case patients isolates are identical and linked.

Facility A: Site Inspection, Infection Control Training, and Review of Procedures

At Facility A, the DON had responsibility for infection control oversight; she had worked at the facility for 3 months and had no formal certification in infection control. The two LPNs conducting fingerstick glucose monitoring began work at the facility in October 2006, during the estimated exposure period. The nurses could not identify any changes in procedure between August 2003, when Facility A opened, and late 2006, when the exposures occurred, that would explain the outbreak.

The facility required nurses to undergo infection control training, including risks for bloodborne pathogen transmission, twice a year and at hire. A nurse from the company supplying the glucose monitoring equipment reportedly gave residents and staff diabetes mellitus education quarterly. This service included training on proper use of glucose monitoring devices. Despite this training, Facility A nurses indicated that they were not aware that HBV could be transmitted from resident to resident during fingerstick blood glucose monitoring.

There were 31 residents with diabetes mellitus at the facility at the time of the investigation. Each morning, they had fasting fingerstick blood glucose monitoring performed at the nursing station and received medication, including insulin if indicated. The nursing staff performed all fingersticks; residents were not allowed to perform their own testing. Nurses were supplied with a pocket-sized alcohol hand gel vial and were instructed to change gloves and use the gel between each fingerstick procedure. A single shared blood glucose meter was used for all residents. There was a written policy to clean the glucometer with alcohol between each use. The nurses used single-use, auto-disabling lancets. Residents requiring insulin received the injection immediately after obtaining their blood glucose monitoring results. Insulin was prepared using multidose vials that were dedicated to individual patients; there were no insulin pens or shared insulin vials in use at the facility.

Facility A: 6-Month Follow-Up

Follow-up testing indicated that at least three new chronic hepatitis B infections resulted from the outbreak at Facility A. In November, 2007, follow-up HBV serological tests were performed on four of the HBsAg-positive residents; three remained HBsAg positive (2 acute cases, 1 probable acute case), whereas one acute case had cleared the infection. The remaining three residents with acute HBV infection had been transferred to nursing homes and were lost to follow-up.

The probable chronic case did not have a known prior history of HBV infection, had normal liver function tests at the time of the initial investigation, and remained HBsAg positive and anti-HBc IgM negative in May 2007, 2 months after the initial testing. This individual moved to a LTCF and was lost to follow-up.

Facility B: Case Identification

During early 2007, as part of the investigation of the initial case, serological testing of other Facility B residents to identify additional asymptomatic cases was not considered. As part of the 2008 investigation at the facility, all of the other residents who were receiving fingerstick blood glucose monitoring at that time were tested, and four residents with asymptomatic acute HBV infections (HBsAg+, IgM+) were identified. None of the six nurses performing fingerstick blood glucose testing had serological evidence of HBV infection.

Review of HBV cases recorded in the state surveillance system identified an additional Facility B resident who had been hospitalized with acute hepatitis B in October 2007. Retrospective record review revealed that this resident received fingerstick blood glucose monitoring.

In all, seven acute HBV cases were identified from Facility B between January 2007 and April 2008 from among residents who received assistance with blood glucose monitoring during this period; three of the infected patients were hospitalized.

Facility B: Site Inspection, Infection Control Training, and Review of Procedures

At Facility B, the DON had responsibility for infection control oversight and had been working for the facility for 3 months; she had no formal certification in infection control. As a result of the turnover in nursing leadership, there was no institutional memory of the initial acute hepatitis B case that occurred in January 2007, the CCDPH investigation conducted at that time, or the infection control recommendations that were given.

Employees were required to receive infection control training annually from the DON. New residents with diabetes mellitus received training on the proper use of blood glucose monitoring devices from facility nursing staff. At admission and at least annually, the ability of each resident with diabetes mellitus to perform glucose monitoring was assessed. The Facility B nurses were not aware that HBV could be transmitted from resident to resident during fingerstick blood glucose monitoring.

The residents themselves or the LPN assigned to the floor performed fingerstick blood glucose monitoring in each resident's room. Each resident with diabetes mellitus had a dedicated glucometer, lancet devices (single use, auto-disabling), other supplies, and sharps containers stored in their room, although the nurses also carried a tray with a spare glucometer and other supplies from room to room when they performed the testing.

Blood glucose monitoring records were incomplete, and it was not possible in many instances to determine from the record review whether a resident had blood glucose monitoring performed, whether the nurse or the resident performed the testing, and which nurse was responsible, but six of the seven acute hepatitis B cases were known to have received blood glucose monitoring assistance from the same two nurses (morning and afternoon shifts).

Nurses were observed to violate Facility B's written infection-prevention protocols as follows. A nurse who cared for the HBV-infected residents worked at multiple facilities and carried a glucometer with her between facilities for use on residents, gloves were not worn while performing fingerstick testing, hands were not washed between residents, the spare glucometer was often used rather than dedicated resident equipment, and the glucometer was not always disinfected between residents.

DISCUSSION

Outbreaks of HBV infection occurred in two ALFs, probably spread by nurses from resident to resident during fingerstick blood glucose monitoring. The HBV DNA sequencing results for Facility A supported the hypothesis of transmission within the facility.

Previous HBV outbreaks in patients with diabetes mellitus reported in the literature were primarily associated with sharing of spring-loaded, penlet fingerstick devices, despite long-standing instructions that these are personal care devices that should not be shared.^1,3,4,13,14 Nevertheless, autodisabling, single-use lancets were used at both of the facilities reported here; HBV contamination of glucose meters, diabetic care supplies, or the hands of nurses was implicated as the source for both outbreaks. Opportunities for blood contamination of these sources has been identified during investigation of other LTCF HBV outbreaks,¹⁵ and a study testing shared glucose monitors in patient care areas at 12 hospitals revealed widespread blood contamination.¹⁶ Recommended practices for preventing patient-to-patient transmission of hepatitis viruses from diabetes care procedures in long-term care settings are summarized in Table 2.

Table 2.   Recommended Practices for Preventing Patient-to-Patient Transmission of Hepatitis Viruses from Procedures for the Care of Diabetes Mellitus in Long-Term Care Settings

Diabetes mellitus care procedures and techniques

• Prepare medications such as insulin in a centralized medication area; multiple-dose insulin vials or insulin pens should be assigned to individual patients and labeled appropriately.

• Never reuse insulin pens, needles, syringes, or lancets.

• Restrict use of fingerstick capillary blood sampling devices to individual patients. Select single-use lancets that permanently retract upon puncture.

• Dispose of used injection equipment, fingerstick devices, and lancets at the point of use in an approved sharps container.

• Environmental surfaces such as glucose meters should be cleaned and disinfected regularly and any time contamination with blood or body fluids occurs or is suspected.

• Glucose meters should be assigned to individual patients. If a glucose meter that has been used for one patient must be reused for another patient, the device must be cleaned and disinfected.

• Unused supplies and medications should be maintained in clean areas separate from used supplies and equipment (e.g., glucose meters). Do not carry supplies and medications in pockets.

Hand hygiene and gloves

• Wear gloves during fingerstick glucose monitoring and during any other procedure that involves potential exposure to blood or body fluids.

• Change gloves between patient contacts. Change gloves that have touched potentially blood-contaminated objects or fingerstick wounds before touching clean surfaces. Discard gloves in appropriate receptacles.

• Perform hand hygiene (hand washing with soap and water or use of an alcohol-based hand rub) immediately after removal of gloves and before touching other medical supplies intended for use on other residents.

Medical management, training, and oversight

• Review regularly individual patients' schedules for fingerstick blood glucose sampling and insulin administration and reduce the number of percutaneous procedures to the minimum necessary for appropriate medical management of diabetes mellitus and its complications.

• Provide a full hepatitis B vaccination series to all previously unvaccinated staff persons whose activities involve contact with blood or body fluids.

• Establish responsibility for oversight of infection control activities. Provide staff members who assume responsibilities involving percutaneous procedures with infection control training. Investigate and report any suspected case that may represent a newly acquired bloodborne infection.

• Assess adherence to infection control recommendations for blood glucose monitoring and insulin administration by periodically observing personnel and tracking use of supplies.

Adapted from Reference 3 and the Centers for Disease Control and Prevention, Blood Glucose Monitoring and the Risk of Viral Hepatitis. Available at http://www.cdc.gov/hepatitis/Populations/GlucoseMonitoring.htm

Delays in diagnosis hampered the investigation of both outbreaks. Healthcare providers should consider acute HBV infection for residents of LTCFs with diabetes mellitus with nonspecific gastrointestinal symptoms.^11,12 An estimated 50% to 70% of adults with acute hepatitis B infection are asymptomatic; infected individuals may not be diagnosed, and they pose an unrecognized risk to others. The probable chronic hepatitis B case believed to be the index case for the Facility A outbreak was not known to be HBV infected, and eight of the 14 acute cases (57%) identified during the two outbreak investigations were asymptomatic at the time of diagnosis.

The nurses at both affected facilities did not recognize that HBV could be spread from resident to resident during fingerstick blood glucose monitoring. Although the nurses had documentation of bloodborne pathogen and infection control training, that knowledge was not applied in practice or did not fully address infection prevention during diabetes mellitus care. Published guidelines exist for infection control and safe injection practices to prevent patient-to-patient, bloodborne pathogen transmission during diabetes mellitus care procedures.^3,17,18 These guidelines include assigning separate glucometers to individual patients, keeping trays or carts with supplies outside of patient rooms, restricting multidose insulin vials to individual patients, wearing gloves during fingerstick glucose monitoring, and changing gloves and performing hand hygiene between patient contacts (Table 2).

Identification of a single LTCF resident with suspected acute HBV infection should trigger a rapid response, involving the local health department, to search for additional cases, evaluate fingerstick blood glucose monitoring practices, and identify other appropriate interventions. In the future, CCDPH will consider obtaining HBV serological testing for all residents of the facility undergoing fingerstick blood glucose monitoring whenever there is a single case of acute HBV infection in this population.

The two hepatitis B outbreaks described were difficult to detect, required considerable resources to respond to, and resulted in significant morbidity. A total of 14 acute HBV infections were identified in ALF residents; six persons were hospitalized. At least three residents did not clear the infection and developed chronic hepatitis B. Strict adherence to aseptic technique by the facility nurses, to prevent contamination of equipment with infected blood during fingerstick glucose monitoring, could have prevented these outbreaks. Medical directors, DONs, and ALF administrators are encouraged to review practices and ensure they do not have conditions in place that would facilitate this type of transmission.

ACKNOWLEDGMENTS

The authors would like to thank Yury Khudyakov, PhD, Head, Division of Viral Hepatitis Molecular Epidemiology Laboratory, CDC, for oversight of the HBV DNA sequence analysis; Craig Conover, MD, State Epidemiologist, Illinois Department of Public Health, for guidance during this outbreak and review of the manuscript; and Nicola Thompson, PhD, MS, Division of Viral Hepatitis, CDC, for review of the manuscript.

Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the authors and has determined that the authors have no financial or any other kind of personal conflicts with this paper.

Author Contributions: Catherine Counard was the lead investigator for the Facility A and Facility B outbreaks and was the lead author. Joseph Perz provided subject matter expertise, interpreted data, and prepared the manuscript. Purisima Linchangco was the investigator of the Facility A and Facility B outbreaks, compiled data from both outbreaks, and took part in manuscript preparation. Demian Christiansen analyzed the data from Facility A and prepared the manuscript. Lilia Ganova-Raeva and Guoliang Xia performed the HBV DNA sequencing analysis and took part in the manuscript preparation. Steven Jones compiled data from the Facility A outbreak and took part in the manuscript preparation. Michael Vernon supervised the Facility A and Facility B investigations, provided subject matter expertise, interpreted data, and took part in manuscript preparation.

Sponsor's Role: None.