Microbial safety implications of in-use topical diagnostic ophthalmic medications in eye clinics in Ghana.

PURPOSE: To determine the microbial contaminants and its clinical importance in topical diagnostic ophthalmic medications (cycloplegics/mydriatics and miotics) in eye clinics in Ghana.
METHOD: A cross-section of eye clinics was sampled for the diagnostic agents (Atropine, Phenylephrine, Tropicamide and Cyclopentolate, Pilocarpine). Standard laboratory procedures and protocols were observed in culturing the samples on different Agars. Microscopy and various biochemical tests were performed to identify microbial species. Antimicrobial susceptibility testing was also performed to ascertain the clinical importance of the isolated microbes.
RESULTS: A total of 113 samples were obtained, from which 334 bacteria were isolated which included Bacilli spp. 91(27.25%), Coagulase Negative Staphylococci spp. 59(17.66%), Moraxella spp. 47(14.07%), Staphylococcus aureus 41(12.27%), Streptococcus spp. 21(6.29%), Klebsiella spp. 20(5.99%), Pseudomonas spp. 13(3.89%), Proteus spp. 12(3.59%), Escherichia coli. 12 (3.59%), Serratia spp. 10(2.99%), Shigella spp. 7(2.09%), Salmonella spp. 1(0.3%). There were 96 isolated fungal contaminants mainly Penicillium spp. 41(42.71%), Cephalosporium spp. 19(19.79%), Cladosporium spp. 15(15.63%), Aspergillus spp. 13(13.54%), Cercospora spp. 8(8.33%). The diagnostic agent with the most bacteria contamination was Phenylephrine 90 (26.95%) and the least being Pilocarpine 49 (14.67%). Also, the diagnostic agent with the most fungal contamination was Cyclopentolate 29 (30.2%) and the least was Tropicamide and Pilocarpine with 15 (15.63%) each. Gentamicin and Ciprofloxacin were the only antibiotics that showed 100% activity against all the bacterial isolates. Fungal contaminants were more susceptible to Ketoconazole as compared to Fluconazole.
CONCLUSION: Topical diagnostic ophthalmic preparations used in clinical settings in Ghana are contaminated with clinically important bacteria and fungi.

Introduction

Infectious diseases remain an important issue globally with Sub-Sahara Africa presenting a high proportion of the global burden. Several factors including poor personal hygiene, contaminated surgical theaters, diagnostic agents and other health care material contribute to the high burden of infectious diseases of which eye infections form a part. Some recalcitrant infectious diseases of the eye such as microbial keratitis and endophthalmitis have been named to contribute to the increased incidence of ocular morbidity and blindness worldwide.

Currently, many researchers have stated that ophthalmic solutions contain pathogenic contaminants associated with ocular infections in clinical settings. A contaminated topical ophthalmic agent serves as a potential cause of avoidable ocular diseases and/or aggravation of pre-existing infections. Apart from the risk of infection, bacterial contamination of ocular diagnostic agent may alter the pH of the solution and therefore reduce the efficacy of the medication. The following has been suggested as the probable factors for this contamination; cross-contamination from diagnostic ophthalmic equipment, use on multiple patients, white coats of clinicians, the duration of use, unsterile hospital environment. Plastic bottles which is the most common container for eye drops have been reported to be more contaminated near the bottle cap. The type and concentration of preservatives in most of these diagnostic agents have been reported to cause disruptions of the ocular surface and this has the potential of enhancing microbial attachment and penetration as well. Despite the intent of preservatives in multiple-use bottles (eye drops meant for multiple usages beyond a day) to prevent contaminations, some microbial contaminations have still been recorded. It is therefore not surprising that most eye infections are attributed to exogenous microbes.

Ocular diagnostic drugs are pharmaceutical agents that are used to test for the normal functioning and impairment of the eye. Thus, they aid in the effective diagnosis of disorders associated with the eyes. These agents are also useful in the examination of the ocular media and fundus, to improve retinal photography and in techniques where anesthesia is required. They are also used in cycloplegic refraction and for the detection of lesions in the visual pathway. These medications are very essential in ophthalmic practice, therefore, should they serve as a means of transmitting infectious pathogens would have public health implications. In response to thiqs problem, there is the need to investigate specific microbial contaminants of these agents so as to inform policy on patient safety and nosocomial infections control. This study was carried out to investigate the prevalence of microbial contamination of diagnostic pharmaceutical agents (DPAs) used in eye clinics in Ghana and to ascertain the clinical importance of these microbes through antimicrobial susceptibility testing. These commonly used diagnostic agents included; cycloplegics and mydriatics i.e. Atropine, Phenylephrine, Tropicamide and Cyclopentolate, miotics i.e. Pilocarpine.

Materials and methods

Sampling procedure

A List of 188 eye clinics was obtained from the Ghana Eye Secretariat of which fifty (50) eye clinics were proportionately selected across the 10 regions of the country. The criterion for the selection was based on what the researchers considered as a major clinic. Thus, only clinics with the full complement of eye care staff and referral centers were selected. These were deemed to perform procedures that would require the use of most of these diagnostics. In each of the selected clinics, the DPAs were collected based on their availability once access was granted to the researchers.

Sample collection procedure

These clinics were not aware they had been selected to participate in the study and were also not given prior notice of our visit to collect samples. This was to avoid any changes and modifications to their normal protocol on the use of these agents. Sterile airtight bottles (one for each sample) wrapped in a pre-sterilized plastic zip bags were used to collect 2 ml each of the samples on the clinicians’ desk meant for routine use. Each sample was drawn with a separate sterile 2.5 ml syringe fitted with a 25 gauge needle. In drawing the samples, a sterile surgical glove was worn to avoid the introduction of hand to sample contamination. Once the samples were collected they were placed into a different labeled sterilized plastic zip bag which was firmly sealed for transportation to the laboratory. All the samples collected had been opened and had been in use for at least 2 weeks and at most 8 weeks. Only one facility had their sample stored at 4 °C, all others were stored at room temperature. Another set of the diagnostic ophthalmic medications as per their availability were obtained from the dispensaries of these facilities to serve as controls for those opened and inuse. The controls were unopened unexpired diagnostic eye drops in multiple use bottles on the shelves (kept at room temperature) of the various dispensary units of the facilities where the samples were collected. They were as well transported under aseptic conditions to the laboratory for the same investigations. All samples were processed within 12 h of collection.

Culturing and identification of microbes

Serial dilution of the sample was done to the fifth dilution using peptone water. The 4th and 5th dilution were plated with nutrient agar using pour plating method and incubated for 18–24 h at 37 °C. Growth on nutrient agar we examined qualitatively and quantitatively and the findings were recorded. Each colony of bacteria isolated from the nutrient agar was inoculated in nutrient broth and incubated overnight under an aerobic condition at 37 °C. Colonies of bacteria growth from nutrient broth were sub-cultured on MacConkey and Blood agar (all from Oxoid Limited, Thermo Fisher Scientific Inc., UK) using streaking method to obtained pure colonies by incubating overnight under an aerobic condition at 37 °C. Loop flaming was always done per plate streaking to avoid contamination of other samples. After incubation, morphological examination of colonies was recorded and further identification was done. Microscopy and Gram staining were done for pure colonies obtain from both Blood and MacConkey agars to differentiate Gram-positive bacteria from Gram-negative bacteria. Gram-positive bacteria obtained were further identified using Catalase, Coagulase and Mannitol agars and gram-negative bacteria were further identified using Oxidase, Indole, Urease, Citrate and Triple sugar agar (TSI). Identified bacteria were subjected to antibiotic sensitivity testing.

Prepared Sabouraud Dextrose Agar, SDA (Oxoid Limited, Thermo Fisher Scientific Inc., UK) in a conical flask and Petri dishes were autoclaved. SDA was allowed to cool to 45–50 °C, and 50 mg of Chloramphenicol was added to 1 L agar and swelled to mix. The SDA media was then poured into Petri dishes to solidify. Diagnostic agents (Phenylephrine, Atropine, Cyclopentolate, Pilocarpine, and Tropicamide) were plated on the prepared SDA using streaking method and incubated at room temperature for 5–7 days under aerobic condition. Morphological characteristics of various fungi growth were examined and recorded. Lactophenol cotton staining preparation was done for each fungus growth, for microscopic examination. Morphological characteristics of stained fungus under the microscope were examined and recorded.

Antimicrobial sensitivity testing

Identified Subculture organisms from both blood and MacConkey agar were picked using sterile loops and emulsified pure colonies in 3 ml of sterile normal saline. Its turbidity was checked by comparing it to 0.5McFarland standard solutions. An adequate lighting was ensured to visually compare the inoculum tube and the 0.5 McFarland standards against a card with a white background and contrasting black line. An already prepared Mueller Hinton agar plates with dried surface were inoculated by streaking the prepared colony in the saline solution by swabbing over its entire surface. Placement of the antibiotics disks (A multi-disk for antimicrobial susceptibility testing from Axiom laboratories) using sterilized forceps aseptically was used to place the antimicrobial impregnated disk on the surface of the agar. The disk was press gently with the help of the forceps to ensure a complete contact with the agar surface and proper diffusion of the drugs into the agar and cover the plate. The plate was inverted and incubated for 24 h at 37 °C. Measurement of inhibition zone size was done from the center of the disk to a point on the circumference of the zone where a distinct edge is to the nearest millimeter.

The method of well diffusion was employed in the fungal sensitivity testing which assessed the efficacy of Ketoconazole and Fluconazole at dose levels of 5–15 mg/ml. A sterile swab was dipped into the standardized inoculum and used to inoculate evenly the surface of already prepared Sabouraud Dextrose Agar. The wells were created in the plated media and then incubated at 25 °C for 5 days.

Statistical analysis

Using one-way ANOVA the mean bacterial contaminations were compared among the regions. One-way ANOVA followed by Bonferroni posthoc test was used to compare the mean values of the various fungi treated with Ketoconazole and Fluconazole. Values were expressed as the mean ± standard error of the mean. p ≤ 0.05 was considered to be statistically significant. GraphPad Prism (version 7; GraphPad, La Jolla, CA, USA) was used in the data analysis.

Ethical considerations

The study protocol was approved by the Department of Biomedical Science Ethics Committee, University of Cape Coast. Permission to access the facilities was sought from their respective managements on arrival of the research team to the facility. Biosafety guidelines for protection of personnel in the laboratory were observed.

Results

Profile of collected samples

Out of the 50 facilities proportionately selected from the list obtained from the Ghana Eye Secretariat, 40 of them granted access and therefore participated in the study. In all 113 samples were obtained across the 10 regions, of which sixteen (16) diagnostic agents were from the Central region, eleven (11) from Brong-Ahafo, twenty-two (22) from Ashanti, eleven (11) from Volta, eleven (11) Greater Accra, five (5) from Eastern, nine (9) from Western, eight (8) from Northern, eight (8) from Upper East, and twelve (12) from Upper West region. Volta and Eastern regions were the regions with the least contamination compared to Ashanti and Central regions (Fig. 1).

Figure 1

A map showing the means ± SEM levels of contamination by region by with blue indicating the region with the least contamination and the red the region with the most contamination. One-way ANOVA (multiple comparisons) indicated Volta and Eastern Region had less contaminations compared to Ashanti and Central Regions. *Indicates p ≤ 0.005).

Of the 113 multi-dose diagnostics agent were sampled, 28 (24.78%) of the agents were Tropicamide and pilocarpine 15 (13.27%) being the one that was collected in a smaller quantity (Table 1). All the medications collected were preserved either with Benzalkonium chloride, Chlorobutanol or Sodium perborate.

Table 1

A table showing the number of topical ophthalmic diagnostic agents sampled from the various regions in Ghana.

Regions	No. of agents obtained	Atropine	Phenylehrine	Cyclopentolate	Pilocarpine	Tropicamide
Central	16	4	4	3	2	3
Ashanti	22	7	5	5	2	3
Brong Ahafo	11	1	2	3	2	3
Northern	8	0	2	2	2	2
Upper East	8	1	2	3	1	1
Western	9	2	2	1	0	4
Upper West	12	2	3	2	3	2
Volta	11	1	2	2	1	5
Greater Accra	11	2	2	2	1	4
Eastern	5	0	3	0	1	1
Total	113	20	27	23	15	28

Profile of cultured samples

A total of 109 diagnostic agents which constitute 96.46% indicated a positive culture with only 4 samples with no contamination. Bacteriological analysis indicated 334 bacterial isolates from 109 diagnostic agents of which the most, 90 bacteria were isolated from Phenylephrine and least 49 from Pilocarpine contaminated (Table 2). The most occurring bacteria were Bacillus spp and the least were Salmonella spp (Table 3). The numbers of Gram-positive and Gram-negative bacteria isolated from the diagnostic agents were also 212 and 112 respectively. On the Sabouraud Dextrose Agar, 96 fungi were isolated of which 16 (16.67%) were from Atropine, 21 (21.87%) from Phenylephrine, 29 (30.2%) from Cyclopentolate, 15 (15.63%) were from Pilocarpine (15.63%) and 15 from Tropicamide. Five different species of fungi were isolated namely, Aspergillus spp., Cladosporium spp., Cephalosporium spp., Penicillium spp., Cercospora spp. The diagnostic agents with the most fungal contamination were Cyclopentolate and the least were Tropicamide and Pilocarpine.

Table 2

The number and percentages of isolated bacteria from the various diagnostic agents.

Diagnostic agents	No. of isolated bacteria	% of isolated bacteria
Atropine	54	16.18
Phenylephrine	90	26.95
Cyclopentolate	71	21.26
Pilocarpine	49	14.67
Tropicamide	70	20.95
Total	334	100

Table 3

A table showing details of the number of bacteria isolated from the various topical ophthalmic diagnostic agents.

Isolated Gram (+) bacteria	Atropine	Phenyleprine	Cyclopentolate	Pilocarpine	Tropicamide	Total
Escherichea coli	2	5	2	2	1	12 (3.59%)
Klebsiella spp.	2	4	5	4	5	20 (5.99%)
Moraxella catarrhalis	9	16	12	4	6	47 (14.07%)
Proteus spp.	1	5	2	2	2	12 (3.59%)
Pseudomonas spp.	2	3	4	1	3	13 (3.89%)
Serratia spp.	2	1	2	1	4	10 (2.99%)
Shigella spp.	3	2	1	1	0	7 (2.09%)
Bacilli spp.	16	28	21	11	15	91 (27.25%)
Cogulaes negative Staphylococcus. spp.	10	15	14	8	12	59 (17.66%)
Staphylococcus aureus	6	5	6	10	14	41 (12.27%)
Streptococcus spp.	1	6	2	5	7	21 (6.29%)
Salmonella spp.	0	0	0	0	1	1 (0.3%)
Total	54	90	71	49	70	334 (100%)

Antimicrobial susceptibility test

Table 4 is a summary of these results for antimicrobial susceptibility test of isolated bacteria for bacteria found.Fig. 2, table 5 and fig. 3, table 6 are also the summaries of the minimum inhibition concentration (MIC) of Ketoconazole and Fluconazole on identified fungi species. Ketoconazole and Fluconazole at doses of 5–15 mg/ml did show a comparable inhibitory effect for all fungal isolates (p > 0.05).

Table 4

The antibiotic susceptibility pattern of the isolated bateria.

Isolated bacteria	AS	BA	CF	TZP	CH	CP	CI	TE	OF	GM	AK	LE	CX	PR	PF	LM	RF
Klebsiella spp.	R	R	R	S	R	S	R	S	S	S	S	S	n/a	n/a	n/a	n/a	n/a
Pseudomona spp.	R	R	I	S	R	S	R	R	S	S	S	S	n/a	n/a	n/a	n/a	n/a
Escherichea coli	R	R	R	I	I	S	R	S	I	S	S	I	n/a	n/a	n/a	n/a	n/a
Moraxella catarrhalis	R	R	I	S	I	S	R	S	S	S	S	R	n/a	n/a	n/a	n/a	n/a
Proteus spp.	R	S	R	S	R	S	R	S	S	S	S	I	n/a	n/a	n/a	n/a	n/a
Shigella spp.	R	S	R	S	R	S	R	R	S	S	S	S	n/a	n/a	n/a	n/a	n/a
Cogulaes negative Staphylococcus spp.	R	S	R	n/a	n/a	S	n/a	I	S	S	n/a	n/a	R	R	I	R	R
Bacillus spp.	R	R	R	n/a	n/a	S	n/a	R	R	S	n/a	n/a	R	R	I	R	R
Staphylococcus aureus	R	R	R	n/a	n/a	S	n/a	R	S	S	n/a	n/a	R	R	R	R	R
Streptococcus spp.	R	R	R	n/a	n/a	S	n/a	S	I	S	n/a	n/a	R	R	R	R	R
Salmonella spp	R	R	R	I	I	S	R	S	I	S	S	S	n/a	n/a	n/a	n/a	n/a

GM: gentamicin, OF: ofloxacin, AK: amikacin, LE: levofloxacin, TE: tetracycline, AS: ansamycin, BA: bacitacin, CF: cephalothin, TZP: tabzobactam/piperacillin, CH: chloramphenicol, LM: lactam; RF: rifampicin, CX: cefotaxime, PR: primaxin.

R – resistant; S – sensitive; I – intermediate; N/A – not applicable.

Figure 2

A plot the area under the curve of the concentrations of Ketoconazole against zones of inhibition. One-way ANOVA indicated that Ketoconazole at all dose levels had a comparable effect on all 5 species isolated (Aspergillus spp., Cladosporium spp., Cephalosporium spp., Penicillium spp., Cercospora spp.).

Table 5

A table showing details of the mean ± SD MICs of Ketaconazole for the fungal isolates.

Fungal isolate	5 mgMean MIC ± SD mm	10 mgMean MIC ± SD mm	15 mgMean MIC ± SD mm
Penicillium spp.	8.267 ± 0.252	11.067 ± 0.153	14.00 ± 0.200
Cephalosporium spp.	16.300 ± 0.300	17.00 ± 0.000	18.267 ± 0.153
Aspergillus spp.	17.00 ± 0.200	19.067 ± 0.058	20.033 ± 0.950
Cercospora spp.	9.067 ± 0.153	10.467 ± 0.208	0.208 ± 0.153
Cladosporium spp.	7.33 ± 0.153	15.633 ± 0.252	21.369 ± 0.643

One-way ANOVA indicated that Ketoconazole at all dose levels had a comparable effect on all 5 species isolated (Aspergillus spp., Cladosporium spp., Cephalosporium spp., Penicillium spp., Cercospora spp.)

MIC: minimum inhibition concentration; SD: standard deviation.

Figure 3

A plot the area under the curve of the concentrations of Fluconazole against zones of inhibition. One-way ANOVA indicated that Fluconazole at all dose levels had a comparable effect on all 5 species isolated (Aspergillus spp., Cladosporium spp., Cephalosporium spp., Penicillium spp., Cercospora spp.).

Table 6

A table showing details of the mean ± SD MICs of Fluconazole for the fungal isolates.

Fungal isolate	5 mgMean (MIC) ± SD mm	10 mgMean (MIC) ± SD mm	15 mgMean (MIC) ± SD mm
Penicillium spp	11.40 ± 0.100	14.467 ± 0.252	15.60 ± 0.200
Cephalosporium spp	10.033 ± 0.058	18.267 ± 0.115	21.000 ± 0.000
Aspergillus spp	18.00 ± 0.100	17.20 ± 0.000	20.167 ± 0.153
Cercospora spp	8.767 ± 0.115	12.033 ± 0.058	18.50 ± 0.100
Cladosporium spp	9.167 ± 0.058	17.330 ± 0.153	19.267 ± 0.058

One-way ANOVA indicated that Ketoconazole at all dose levels had a comparable effect on all 5 species isolated (Aspergillus spp., Cladosporium spp., Cephalosporium spp., Penicillium spp., Cercospora spp.).

MIC: minimum inhibition concentration; SD: standard deviation.

Discussion

It is unfortunate to visit the clinic to seek remedy and to return home with a different disorder or an aggravation of your existing condition. Such is the experience of many who visit healthcare facilities especially in low resource settings of the world. It has been established that contaminated ophthalmic solutions could be an important conduit for such infections.10, 11 More so, the multidose form which is impregnated with preservatives such as benzalkonium chloride is known to alter ocular surface integrity thereby aiding the penetration of the drug and its accompanying contaminants. From literature, Chlorobutanol also alters the normal cellular movement and proliferative activity resulting in conjunctival and corneal cytotoxicity. With long-term used of ocular agents containing Chlorobutanol as a preservative, significant keratitis has been noted due to degeneration of human corneal epithelial cells. These alterations caused by Chlorobutanol to the ocular surface could rather fester the invasion by microbes in cases where contaminated solutions are used.

The contaminated topical diagnostic preparations could have resulted from the clinician's hand contact with the nozzle while opening for usage, or it comes in contact with ocular tissue, multi-user handling, duration of use and other factor involving environmental dynamics. Caps of these agents usually harbors airborne microbes which may contaminate these agents. These factors have the potential to overwhelm the agents meant for preservation due to their limited concentrations in these eye drops rendering them ineffective. The extent of contamination recorded in this study far outweighs those recorded in other several studies involving ophthalmic solutions.15, 16, 17, 18, 19 None of the regions meet the strict sterility criteria required of ophthalmic solutions despite a statistically significant lower level of contaminations recorded for Volta and Eastern Regions relative to Ashanti and Central regions (p ≤ 0.05). The microbes found in this study further augment the possible causes of the contamination (use on multiple patients, white coats of clinicians, the duration of use, unsterile hospital environment, etc.) as most of the isolated microbes were either associated with normal flora of humans (from the skin e.g. Streptococcus spp., nasopharynx e.g. Moraxella catarrhalis and the ocular tissues e.g. Staphylococcus aureus and Streptococcus spp.), gastrointestinal tract pathogens e.g. Escherichia coli, and environmental factors by airborne e.g. Penicillium spp, Aspergillus spp etc. Bacteria isolates from the Enterobacteriaceae family such as E. coli., Serratia spp., Shigella spp., Proteus spp. indicates possible fecal contamination of these diagnostic agents. Klebsiella spp which are microbiota found in the nasopharyngeal and gastrointestinal has been reported to cause endophthalmitis. Pseudomonas spp. a common microbial associate of ocular infections are even capable of traversing uncompromised corneal epithelium.

Previous studies have suggested white coats, hospital objects, clinicians mobile, bags and wallet in the transmission cycle of pathogenic organisms in healthcare facilities. Streptococcus spp and M. catarrhalis which are nasopharynx commensal bacteria may contaminate topical diagnostic agent by talking and breathing. Other Staphylococcus spp (Coagulase-negative Staphylococcus spp.) and Streptococcus spp which are normal flora of the skin are indicative of contamination from human contact mainly due to poor handling. The predominance of Bacilli spp in the diagnostic agents is because of the high survival rate of this spore-forming bacteria even in preserved ophthalmic solutions. The constant introduction of non-native bacteria such as Serratia spp, Klebsiella spp, Salmonella spp, and Shigella spp into the eye may have future consequences for the emergence of recalcitrant ocular infections.

The incidence of airborne fungal contamination varied widely, depending on the season, temperature and relative humidity. According to literature, fungal contaminations to diagnostic agents are largely due to airborne flora with Aspergillus spp, Penicillium spp, Cephalosporium spp, and Cladosporium spp.

This suggests that contamination of topical diagnostic preparations was airborne due to geographical, environmental factors and unhygienic clinical settings. The most recalcitrant form of ocular surface infections are of fungal etiologies and could have dire consequence for patients who are innocently infected. This is even much more so; since patients infected with these conditions may present with an unrelated history of contact with vegetative matter leading to misdiagnosis or delayed detection. Penicillium spp has been reported to be a common cause of conjunctival infections. Cephalosporium spp does not only cause scleral infections but affects corneal sensitivity. Cladosporium spp, causes Cladosporium Keratitis and Cercosporium spp is also a known cause of ocular infections.

Ciprofloxacin and Gentamicin showed high activity against all bacteria isolates but the organisms proved resistant against most of the conventional antibiotics. However, all the fungal isolates were susceptible to Fluconazole and Ketoconazole.

Conclusion and recommendations

Conclusion

Topical diagnostic ophthalmic preparations (Atropine, Phenylephrine, Cyclopentolate, Pilocarpine, and Tropicamide) used in clinical settings in Ghana were contaminated with clinically important microbes. These clinically important bacteria isolated are pathogenic and would cause serious secondary eye infection if they come in contact with the eye of immunocompromised persons. Ciprofloxacin and Gentamicin were active against Gram-negative and positive bacteria isolated but most bacteria were resistant or intermediate to multidisc antibiotic susceptibility testing used (from Axiom laboratories). The isolated fungi were susceptible to Ketoconazole and Fluconazole at concentrations of minimum concentration of 5 mg/dl.

Disclosure of funding

This study was solely funded by the authors.

Conflicts of interest

The authors have no conflicts of interest to declare..