Bacterial contamination of multiple-dose

vials: A prevalence study

Frauke Mattner, MD, and Petra Gastmeier, MD
Hannover, Germany

Background: Two patients died of a meningitis caused by Pseudomonas aeruginosa in a hospital in Germany in July 2001, their
infections having been caused by a contaminated contrast media (iomeprol [Imeron]) used as a multiple-dose vial (MDV) over 8
days. Therefore, a prevalence study was performed to investigate the use and contamination of multiple-use vials in a tertiary
hospital.

Methods: In a 1300-bed hospital on a specific day in November 2001, all used MDVs were collected by the infection control nurses.
Information was recorded about the medication, labeling of vials, storing temperature, wards, and dates of opening. Each vial was
also tested for sterility.

Results: Opened vials were to be found in all wards. Of the 227 vials available, 1 vial and 1 spike were contaminated with
Staphylococcus epidermidis (contamination rate 0.9%; 95% CI, 0.3-2.1). The opening dates were marked on only 114 (50%) MDVs,
15 (13%) of which had already expired. Only 44 (19%) MDVs had been stored in the refrigerator, whereas 109 MDVs contained
medications without any preserving agent.

Conclusion: Results revealed somewhat risky handling of MDVs. In light of a possible high risk in this hospital of about 1
contaminated MDV per day, and in view of many reported outbreaks induced by contaminated MDVs, the following infection
control measures were encouraged: alcohol hand hygiene, the disinfection of gums, observance of the manufacturer’s
recommendations, appropriate storing temperatures, marking the opening time, and avoiding the multiple use of medications not
containing preserving agents. (Am J Infect Control 2004;32:12-6.)

Studies of multiple-dose vials (MDVs) have revealed
considerable variation in bacterial contamination rates,
ranging from 0% to 27%.1 Some of these studies
showed alarming contamination rates,2 whereas others
mentioned only sterile vials, thus favoring the mul-
tiple use of vials for reducing costs and general con-
venience.1,3
Worthy of note are severe iatrogenic infections that
from time to time have been cited in scientific litera-
ture as outbreaks but are in fact caused by the use
of bacterially contaminated multiple-use vials. Often
these outbreaks are reported in the newspapers. A
recent report of 2 deaths, after the injection of a
Pseudomonas aeruginosa–contaminated contrast me-
dia (iomeprol [Imeron, Altana, Pharma, Germany]) vial,
was the starting point of our current investigation.4,5
The vial in question was used for 41 patients over
From the Institute for Medical Microbiology and Hospital Epidemiology,
Medical School Hannover.
Reprint requests: Frauke Mattner, MD, Medizinische Hochschule
Hannover Institut für Medizinische Mikrobiologie und Krankenhaush-
ygiene, Carl-Neuberg-Str. 1, 30625 Hannover, Germany.
0196-6553/$30.00
Copyright ª 2004 by the Association for Professionals in Infection
Control and Epidemiology, Inc.
doi:10.1016/j.ajic.2003.06.004
a period of 8 days, iomeprol having been injected
intrathecally by an orthopedist. Two patients sub-
sequently developed meningitis whose strains were
genetically identical to those found in the vial.
One must, therefore, suspect that despite various
scientific publications about vial contamination over
the last 40 years, infection control behavior seems to
have remained unchanged to the present day. To gain
a current overview of the risk of MDVs in a 1300-bed
tertiary hospital, a 1-point prevalence study was
performed.
METHODS
MDVs—multiple-use vials
The term multiple-dose vial is used widely and in
a confusing manner. In daily practice, the term
designates any kind of vial that may have been used
more than once and kept for potential reuse. A ‘‘true’’
MDV describes a vial in which antibacterial preserva-
tives are present and which may be used more than
once and following the manufacturer’s recommenda-
tions (eg, insulin, heparin, and octeotride).6 Single-
dose vials are intended to be used only once (eg,
salines, antibiotics, anesthetics, etc). All opened con-
tainers that had been used at least once and kept for
potential reuse (thus the term multiple-use vial) were
investigated in this study.

12
 Mattner and Gastmeier February 2004 13

Table 1. Distribution of MDVs Table 2. Multiple-use vials and type of medication

Expired MDVs* Number of vials
Collected Dated (from dated Medication type (%) collected
Type of unit MDVs MDVs (%) MDVs only)
Salines 0.9% 77 (34%)
Intensive care 109 51 (47%) 11 Heparins 41 (18%)
Medical 35 15 (43%) 1 Glucose 16 (7%)
Pediatric 26 24 (93%) 0 Water for injections 16 (7%)
Radiology 2 0 (0%) 0 Insulin 32 (14%)
Surgical 39 18 (46%) 2 Miscellaneous 45 (20%)
Neurology 5 4 (80%) 1
Ambulances 11 2 (18%) 0
Total 227 114 (50%) 15
ampoules and 18 (8%) were syringes prepared by
MDVs, Multiple-dose vials.
A vial was defined as ‘‘expired’’ if it contained no preservatives and had been
nurses for flushing intravascular lines.
collected more than 24 hours after its opening date. A total of 177 (77.9%) MDVs was collected in nursing
rooms; none of them were marked with any patient’s
name, which indicated that they were probably not
Collecting the vials used for only 1 patient. Of these, 33 (14.5%) were
Infection control nurses, without prior warning, collected in patients’ rooms, the rest elsewhere. For 84
collected all the opened vials of 1 day from every ward. (37%) MDVs, mini-spikes were used; for 10 (4%),
The following data were recorded for each vial: medi- canulas. Intensive care units (105 beds; 1.04 MDVs/bed)
cation, location, labeling, or nonlabeling of the vials, used 10 times more MDVs than did other wards (1169
the date and time of opening, and storing temperature. beds; 0.1 MDV/bed). Table 1 shows the number of MDVs
The use of spikes containing bacterial filters was also from different wards.
noted.
A vial was deemed to have expired if it contained no Labeling of vials
preservatives and had been collected more than 24 Of 227 MDVs, 113 (50%) were undated. On 2 MDVs,
hours after its opening date, as was recommended no medication type was indicated, and on 7 vials
many years ago. containing an individual concentration of a certain
medication, the concentration was not given. Of 114
Laboratory diagnostic (50%) dated vials, 15 (13%) had expired up to 14 days
before (median = 1 day).
Each vial was tested under aseptic conditions on
reaching the laboratory. The gums of the vials were Medication and storing temperature
swabbed with alcohol before any of the content was Medication types are shown in Table 2. There were
removed under laminar air flow conditions. Contents various miscellaneous medications such as catechol-
exceeding 10 mL were filtered and the filters then amines, atropine, antibiotics, midazolam, furosemide,
placed onto a blood agar plate. Plates were cultured for potassium chloride, iopromide, and some others. About
48 hours at 378C, then 1 mL of the substance was put 50% of all MDVs contained preservative-free medica-
into a tube containing thioglycolate broth and cultured tions such as glucose 5%, saline 0.9%, and water for
at 378C for 1 week. Contents of less than 10 mL were injections. Thirty-seven (16%) were duplicates, mean-
not filtered, but 2 mL were placed directly onto blood ing that more than 1 open vial with the same type of
agar plates. Medications containing preservatives medication existed on the same ward. A total of 183
were inactivated with caso-bouillon (3% tween 80; (81%) vials had been stored at room temperature, the
3% saponin; 0.1% histidin; 0.1% cystein). Blood agar rest in a refrigerator at 48C.
plates and thioglycolate broth were examined for
bacterial growth, the identification of bacteria being Bacterial contamination
carried out according to routine diagnostic procedures. From 1 vial containing a saline solution, Staphylo-
Mini-spikes (Mini-Spike Plus, B. Braun, Melsungen, coccus epidermidis was cultured. Rinsing broth from
Germany) were taken out of the bottle and flushed with a spike (pulled out of a bottle containing saline
5 mL thioglycolate broth. The broth was then cultured, solution) was also contaminated with S epidermidis.
and bacteria were identified as just described. This led to a contamination prevalence of 0.9% (CI95
0.15%-3%). Both vials had been collected from 2
RESULTS different intensive care units and had been stored at
A total of 227 MDVs was collected from 47 wards room temperature, 1 vial having been opened the
(4.8 MDV/ward). Of these, 209 (92%) were bottles and previous day, the other the day before that.
14 Vol. 32 No. 1 Mattner and Gastmeier

Table 3. Outbreaks caused by multiple-dose vials from 1983 through 2002

Evidence for Year,
Medication Pathogen Infection Patients Death relatedness reference

Propofol E cloacae BSI 4 2 Epidemiologic, molecular biological 200212

Propofol S marcescens BSI, wound 7 2 Epidemiologic, molecular biological 200113
infection
Propofol S aureus Wound 5 Molecular biological 199014
infection
Propofol S aureus BSI 5 Epidemiologic, molecular biological 199715
Propofol S aureus, C albicans, Moraxella Nosocomial 62 Epidemiologic, molecular biological 199516
osloensis, E agglomerans, infections
S marcescens
Propofol K pneumoniae BSI 4 Microbiological 199417
Propofol HCV Hepatitis C 5 Epidemiologic, molecular biological 200118
Saline HCV Hepatitis C 17 Epidemiologic, molecular biological 200319
Saline A xylosoxidans 7 Microbiologic 200020
Saline HCV Hepatitis 2 Molecular biological 200221
Saline and S aureus BSI, Abscess 7 Molecular biological 200022
anesthetic
Saline HBV Hepatitis 30 Molecular biological 200023
Epoetin alfa S liquefaciens BSI 15 Molecular biological 200124
Vaccine DT Streptococci, group A Abscess 2 Microbiologic 19858
Vaccine DTP Streptococci, group A Abscess 9 Microbiologic 199325
Heparin B cepacia BSI 14 Epidemiologic molecular biological 199326
Heparin Plasmodium falciparum Malaria 1 Epidemiologic 200027
Heparin and HCV Hepatitis 11 Epidemiologic molecular biological 200228
saline solution
Dextrose (NICU) E cloacae and P aeruginosa BSI 6 2 Epidemiologic molecular biological 199829
TPN C albicans BSI 8 3 Molecular biological 200030
Anesthetic; HBV Hepatitis 2 Epidemiologic, molecular biological 199931
not specified
Distilled water P pickettii BSI 7 Microbiologic 199132
Bupivacaine HBV Hepatitis 10 Epidemiologic, serotyping 198333
Methyl-prednisone S marcescens Arthritis 10 Microbiologic 198734
Iomeprol P aeruginosa Meningitis 5 2 Molecular biological 20014
Not specified HIV AIDS 1 Epidemiologic, molecular biological 199935
Not specified HCV Hepatitis 11 Epidemiologic, molecular biological 200236
BSI, Bloodstream infection; DT, diphtheria and tetanus toxoids; DTP, diphtheria and tetanus toxoids and pertussis; HBV, hepatitis B virus; HCV, hepatitis C virus; NICU, neonatal
intensive care unit; TPN, total parenteral nutrition.

DISCUSSION aliquots of 100 mL from each MDV for the growing of

Results demonstrate the risk related to handling of
MDVs. Almost any kind of medication was used for
multiple-use vials. Cheap, preservative-free solutions,
such as sterile saline and glucose, were used as MDVs;
the potential contamination risk, possibly leading to
severe infections in patients, had not been realized
at all.
Contamination studies with inconsistent data have
been presented for decades. In 1984, Longfield1 studied
1223 samples from 864 MDVs and could not find any
contamination, whereas the study reviewed other
contamination studies with an overall contamination
rate of 0.6% (24 contaminated vials out of 4036
samples). Ten years later, Melnyk7 reported a contam-
ination rate of 1.4% (1 out of 69). These differences in
contamination rates of MDV may partly be attributed to
methodical details. Thus, Melnyk examined enrich-
ments of solutions, whereas Longfield studied only
bacterial cultures. Melnyk also tested for blood cell
contamination, which indicated the possible risk of
viral transmission in cases of positive contamination.
Also, the types of collected medication are important
for explaining inconsistent contamination rates. Long-
field and Melnyk mainly collected medications that
contained preservatives. Our own data revealed that
almost half of all collected vials contained preservative-
free medications such as sterile salines, glucose, and
water for injections.
The discovery of the 2 cultures positive of S epider-
midis possibly points to a situation allowing cross-
contamination. It is conceivable that many bacterial
species could have been introduced into these vials
and had the potential to cause serious bloodstream
infections.
Although MDV contamination studies have revealed
rather differing contamination rates and found various
 Mattner and Gastmeier
bacterial species, the dramatic consequences of such
contaminations have led again and again to sensational
reports of outbreaks caused by MDVs (Table 3).
At least 17 studies reported outbreaks with various
species of fungi and bacteria, which grew in many
different types of medication, the contaminated vials
having been reused in other patients. A further 9
studies reported transmissions of viruses and 1 trans-
mission of Plasmodium falciparum. Interestingly, no
outbreaks of S epidermidis were mentioned, although
in our study S epidermidis was cultured. A reason for
this apparent discrepancy may lie in the difficulty of
determining this bacterial species as the cause of an
outbreak. Cross-transmissions usually occur only in
small clusters. Probably physicians normally take
S epidermidis to be the cause of catheter-related blood-
stream infection without suspecting transmission of
bacteria from any sources other than the residual flora
on the patient’s skin.
The outbreak data and our own data revealed that
almost every type of medication was subjected to
multiple use. It should be noted here that some
medication types are more likely to cause outbreaks
than others. For instance, medicaments containing
lipids, such as propofol (7 reported outbreaks), seem
to be the most dangerous, followed by preservative-
free solutions. Preservatives in vaccine MDVs do not
prevent short-term bacterial contamination,8 whereas
it had been proven that preserved insulins and
heparins are safe for long periods.9 The use of bacterial
filters containing mini-spikes might appear to remove
the risk of contamination. In our study, though, it
was shown that even spikes could be contaminated.
Consequently, the presence of mini-spikes provides
staff with a false sense of security and leads them to
improper handling of the vials. Using mini-spikes,
therefore, could present additional risk.
Further, the outbreaks reported may represent only
the tip of an iceberg of many outbreaks whose origins
remain unknown. Thus the danger from MDVs cannot
be overestimated. Nevertheless, there are authors who
promote the multiple use of vials on the basis of greater
cost-effectiveness. Kamishira3 reported no bacterial
contamination in 28 vials containing MR (Magnet-
resonance)-contrast media vials. With a theoretical
overall contamination rate of between 0.6% and
1.6%,1 KamishiraÕs study sample was certainly too
small to find any positive vial. On the other hand, the
Monti10 study reported an enormous risk potential
from MDVs and concluded that the cost-effectiveness
of using MDVs was nonexistent if estimated cases of
nosocomial infections were also taken into account.
It may be concluded that transmission risk from
MDVs depends on the extent of the infection con-
trol measures taken by staff and on the use of
February 2004 15
medications containing lipids or preservative-free
medications.
An analysis of this study’s data was presented to the
units, after which it was decided to reemphasize the
infection control measures put forth by the Healthcare
Infection Control Practices Advisory Committee Guide-
lines for the Prevention of Intravascular Catheter-
Related Infections:11 alcohol hand hygiene before any
contact with intravenous medications, the disinfection
of vial gums, observance of storing conditions, and vial
dating. However, the most important recommendation
is to avoid any multiple use of preservative-free vials.
We thank Ilka Fröse and Dagmar Stresemann for their technical assistance and
Undine Baum, Juliane Bitsch, Dagmar Rotermund-Rauchenberger, and Uta Schoener-
Fruehling for collecting all containers.
References
1. Longfield R, Longfield J, Smith LP, Hyams KC, Strohmer ME. Multidose
medication vial sterility: an in-use study and a review of the literature.
Infect Control 1984;5:165-9.
2. Bothe J. Study shows contamination in multiple-dose vials. AORN J
1973;17:111-4.
3. Kamishima SM, Awaya H, Abraham D. Utilization of ‘‘used’’ vials: cost-
effective technique for MR arthrography. J Magn Reson Imaging
2000;12:953-5.
4. Meningitis: Kontrastmittel war infiziert. Hamburger Abendblatt
2001;26:07.
5. Meningitis-Skandal: Zwei Ärzte entlassen. Hamburger Abendblatt
2001;24:08.
6. Dade J, Wilcox M, Kay L. Hazards of multiple use of pharmaceutical
solutions. Lancet 2000;356:1684-5.
7. Melnyk P, Shevchuk YM, Conly JM, Richardson CJ. Contamination
study of multiple-dose vials. Ann Pharmacother 1993;27:274-8.
8. Stetler HC, Garbe PL, Dwyer DM, Facklam RR, Orenstein WA, West
GR, et al. Outbreaks of group: A streptococcal abscesses following
diphtheria-tetanus toxoid-pertussis vaccination. Pediatrics 1985;75:
299-303.
9. Tarr BD, Campbell RK, Workman TM. Stability and sterility of
biosynthetic human insulin stored in plastic insulin syringes for 28 days.
Am J Hosp Pharm 1991;48:2631-4.
10. Monti EJ. The safe use of disposable syringes in anesthesia: cost
effective or costly? CRNA 1995;6:86-90.
11. O’Grady NP, Alexander M, Dellinger EP, Gerberding JL, Heard SO,
Maki DG, et al. Guidelines for the prevention of intravascular catheter-
related infections. Centers for Disease Control and Prevention.
MMWR Recomm Rep 2002;51(RR-10):1-29.
12. Weist K, Wilbrandt B, Herm T, Halle E, Melzer C, Rüden H. Severe
cases of sepsis in an outpatient clinic causes by contaminated
intravenous propofol. Wissenschaftliches programm des 54. DGHM-
Tagung Heidelberg 2002 (Abstract).
13. Henry B, Plante-Jenkins C, Ostrowska K. An outbreak of Serratia
marcescens associated with the anesthetic agent propofol. Am J Infect
Control 2001;29:312-5.
14. Postsurgical infections associated with an extrinsically contaminated
intravenous anesthetic agent—California, Illinois, Maine, and Michigan,
1990. MMWR Morb Mortal Wkly Rep 1990;39: 426-7, 433.
15. Kuehnert MJ, Webb RM, Jochimsen EM, Hancock GA, Arduino MJ,
Hand S, et al. Staphylococcus aureus bloodstream infections among
patients undergoing electroconvulsive therapy traced to breaks in
infection control and possible extrinsic contamination by propofol.
Anesth Analg 1997;85:420-5.
16 Vol. 32 No. 1
16. Bennett SN, McNeil MM, Bland LA, Arduino MJ, Villarino ME, Perrotta
DM, et al. Postoperative infections traced to contamination of an
intravenous anesthetic, propofol. N Engl J Med 1995;333:147-54.
17. Veber B, Gachot B, Bedos JP, Wolff M. Severe sepsis after intravenous
injection of contaminated propofol. Anesthesiology 1994;80:712-3.
18. Massari M, Petrosillo N, Ippolito G, Solforosi L, Bonazzi L, Clementi M,
et al. Transmission of hepatitis C virus in a gynecological surgery
setting. J Clin Microbiol 2001;39:2860-3.
19. Dumpis U, Kovalova Z, Jansons J, Cupane L, Sominskaya I, Michailova
M, et al. An outbreak of HBV and HCV infection in a paediatric
oncology ward: epidemiological investigations and prevention of
further spread. J Med Virol 2003;69:331-8.
20. Granowitz EV, Keenholtz SL. A pseudoepidemic of Alcaligenes
xylosoxidans attributable to contaminated saline. Am J Infect Control
1998;26:146-8.
21. Lagging LM, Aneman C, Nenonen N, Brandberg A, Grip L, Norkrans
G, et al. Nosocomial transmission of HCV in a cardiology ward during
the window phase of infection: an epidemiological and molecular
investigation. Scand J Infect Dis 2002;34:580-2.
22. Michels. Nosokomiale S. aureus Infektionen: Infektionen ausgehend
von kontaminierter Injektionsflüssigkeit. Epidemiologisches Bulletin
2000;10:80.
23. Webster GJ, Hallett R, Whalley SA, Farrington CP, Sharma S, Hamilton
G, et al. Molecular epidemiology of a large outbreak of hepatitis B
linked to autohaemotherapy. Lancet 2000;356:379-84.
24. Grohskopf LA, Roth VR, Feikin DR, Arduino MJ, Carson LA, Tokars JI,
et al. Serratia liquefaciens bloodstream infections from contamination of
epoetin alfa at a hemodialysis center. N Engl J Med 2001;344:1491-7.
25. Simon PA, Chen RT, Elliott JA, Schwartz B. Outbreak of pyogenic
abscesses after diphtheria and tetanus toxoids and pertussis vacci-
nation. Pediatr Infect Dis J 1993;12:368-71.
26. Pegues DA, Carson LA, Anderson RL, Norgard MJ, Argent TA, Jarvis
WR, et al. Outbreak of Pseudomonas cepacia bacteremia in oncology
patients. Clin Infect Dis 1993;16:407-11.
Mattner and Gastmeier
27. Al-Saigul AM, Fontaine RE, Haddad Q. Nosocomial malaria from
contamination of a multidose heparin container with blood. Infect
Control Hosp Epidemiol 2000;21:329-30.
28. Kokubo S, Horii T, Yonekawa O, Ozawa N, Mukaide M. A
phylogenetic-tree analysis elucidating nosocomial transmission of
hepatitis C virus in a haemodialysis unit. J Viral Hepat 2002;9:450-4.
29. Archibald LK, Ramos M, Arduino MJ, Arduino MJ, Aguero SM, Deseda
C, et al. Enterobacter cloacae and Pseudomonas aeruginosa polymicro-
bial bloodstream infections traced to extrinsic contamination of
a dextrose multidose vial. J Pediatr 1998;133:640-4.
30. Tresoldi AT, Padoveze MC, Trabasso P, Veiga JF, Marba ST, von
Nowakonski A, et al. Enterobacter cloacae sepsis outbreak in a
newborn unit caused by contaminated total parenteral nutrition
solution. Am J Infect Control 2000;28:258-61.
31. Kidd-Ljunggren K, Broman E, Ekvall H, Gustavsson O. Nosocomial
transmission of hepatitis B virus infection through multiple-dose vials.
J Hosp Infect 1999;43:57-62.
32. Lacey S, Want SV. Pseudomonas pickettii infections in a paediatric
oncology unit. J Hosp Infect 1991;17:45-51.
33. Alter MJ, Ahtone J, Maynard JE. Hepatitis B virus transmission
associated with a multiple-dose vial in a hemodialysis unit. Ann Intern
Med 1983;99:330-3.
34. Nakashima AK, McCarthy MA, Martone WJ, Anderson RL. Epidemic
septic arthritis caused by Serratia marcescens and associated with
a benzalkonium chloride antiseptic. J Clin Microbiol 1987;25:
1014-8.
35. Katzenstein T, Jorgensen LB, Permin H, Hansen J, Nielsen C, Machuca
R, Gerstoft J. Nosocomial HIV-transmission in an outpatient clinic
detected by epidemiological and phylogenetic analyses. AIDS 1999;
13:1737-44.
36. Silini E, Locasciulli A, Santoleri L, Magliano E, Nosari A, Morra E, et al.
Hepatitis C virus infection in a hematology ward: evidence for
nosocomial transmission and impact on hematologic disease outcome.
Haematologica 2002;87:1200-8.