The importance of comorbidity in the presence of low culture positivity in a education and research hospital in Istanbul

Aims: To assess the clinical outcomes of patients with endophthalmitis.We also aimed the effect of comorbidities on hospitalization time, treatment management and vision gain. Metods: This retrospective study includes 40 eyes of 40 patients.Endophthalmitis was divided into as exogenous and endogenous groups.We investigated that culture results, comorbidities, hospitalization times, treatment management and vision gain.Patients with diabetes mellitus(DM) and/or hypertension(HT) were placed under comorbidity group 1.Patients with chronic distant organ inammation in addition to DM and/or HT were placed under comorbidity group 2. Results: Endophthalmitis were exogenous origin in 25 eyes, endogenous origin in 15 eyes.Culture positivity rate,pars plana vitrectomy(PPV) rate,Intravitreal injection(IVI) rate,rePPV rate,two IVIs rate,more than two IVIs rate in exogenous and endogenous groups were 16%, 20%, respectively(culture positivity); 56%, 86.7%, respectively(PPV); 44%, 13.3%, respectively(IVI); 4%, 20%, respectively(rePPV); 52%, 6.7%, respectively(two IVIs); 20%, 66.7%, respectively(more than two IVIs).In exogenous group, mean hospitalization time in comorbidity group 1 and comorbidity group 2 was 7.3 ± 2.5, 10.5 ± 2.8 days, respectively.Mean number of IVIs in comorbidity group 2 was 28% more than comorbidity group 1.In endogenous group, mean hospitalization time in comorbidity group 1 and comorbidity group 2 was 14.3 ± 10.1, 22.0 ± 6.79 days, respectively.Mean number of IVIs in comorbidity group 2 was 86% more than comorbidity group 1. Conclusions: Because our culture and antibiogram ndings were low, we had to increase surgical procedures and repeat IVI.This study showed that patients with


Introduction
Endophthalmitis is a sight-threatening in ammation of intraocular spaces. It is classi ed as exogenous or endogenous depending on the route of infection. 1,2 Endogenous endophthalmitis is less common than the other type and occurs in 2-40% of all endophthalmitis cases. 3,4 Culture positivity is de ned as the identi cation of a causative agent and the application of the antimicrobial agent to which the causative agent is most sensitive. A precise microbiological diagnosis allows the accurate treatment of endophthalmitis. However, microbiological methods have limitations in detecting pathogens, and clinicians may be left with negative results.
Comorbidity is a disease or condition that coexists with but often is independent of another disease.
Previous studies have demonstrated that the presence of comorbidity has been associated with oxidative stress/in ammation and oxidative stress/in ammation leads to ocular, systemic, and psychiatric diseases or exacerbation of diseases. [5][6][7] In this study, we aimed to investigate the culture results, comorbidities, hospitalization times, treatment management, and vision gains of patients with endogenous and exogenous endophthalmitis. In addition, we investigated the effect of culture results and comorbidities on hospitalization time, treatment management, and vision gain.

Materials And Methods
Between March 2016 and January 2021, 40 eyes of 40 patients who diagnosed endophthalmitis were retrospectively analyzed in ophthalmology clinic of Istanbul Education and Research Hospital. This retrospective, single-center, case series was approved by Istanbul Education and Research Hospital ethics committee. The study adhered to the tenets of the Declaration of Helsinki.
Cases diagnosed with endophthalmitis were identi ed. The patients' charts were subsequently reviewed.
The culture results, systemic comorbidities, hospitalization times, and endophthalmitis management of the patients were determined from their medical records. The patients' initial and nal visual acuity (VA) values were recorded using a decimal VA card. Decimal VA values were converted to the logarithm of the minimum angle of resolution (logMAR) VA values. As established by prior studies, 8,9 the vision levels of hand motion, light perception, and no light perception were assigned with the VA values of 0.5/200, 0.25/200, and 0.125/200, which have logMAR values equivalent to 2.6, 2.9, and 3.2, respectively. Vision gain was calculated by removing initial VA from nal VA as logMAR.
Totally 40 eyes were divided into two groups according to origin as exogenous endophthalmitis (Group 1; n:25) and endogenous endophthalmitis (Group 2; n:15).

Inclusion and exclusion criteria
Patients with exogenous and endogenous endophthalmitis were included in this study. Cases of traumatic endophthalmitis and those associated with scleritis or keratitis were excluded. Endophthalmitis was de ned as a condition where patients presented with a clinical suspicion that was high enough to warrant a vitreous/aqueous tap and PPV and/or IVI. In general, these patients presented with decreased VA values and pain and had signs of intraocular in ammation on examination (generally ≥ 2+ anterior segment cellular reaction and/or posterior segment vitritis). Patients who did not have a vitreous/aqueous culture performed or who were treated with topical steroids without additional interventions were excluded.

Prophylaxis for intravitreal injections and other intraocular surgeries
In all surgical procedures, povidone iodine was routinely used as an antiseptic on skin and ocular surfaces in 10% and 5% concentrations, respectively. Intracameral cefuroxime (1 mg/0.1 ml) was applied at the end of cataract surgery, and topical moxi oxacin phthalmic solution 0.5% was administered postoperatively. Subconjunctival gentamicin (0.4 ml, 20 mg/ml) and cefazoline (0.4 ml, 100 mg/ml) injections were applied at the end of PPV surgery, and topical moxi oxacin phthalmic solution 0.5% and oral moxi oxacin (400 mg) was used postoperatively. All intravitreal injections were performed in an operating room. Injections were performed with 30-gauge needles 3.5 mm from the corneoscleral limbus. Commercially available topical oxytetracycline (5 mg) combined with polymyxin B pomade (10.000 U) was used at the end of intravitreal injections. Topical moxi oxacin phthalmic solution 0.5% were applied after intravitreal injections.

Microbiological techniques
Intraocular samples were collected in the operating room from patients under anesthesia. The samples included aqueous and vitreous humors obtained by vitreous taps or vitreous biopsies. The samples were processed within half an hour by rst inoculating them onto culture media and then performing a direct smear examination of Gram-stained samples.
The collected samples were placed in Fluid thioglycolate medium, Blood agar, Chocolate agar, Eosin methylene blue agar, and Sabouraud dextrose agar in the operating room. Gram stain results from vitreous/aqueous cultures were available within 1 day after the taps, and culture plates were generally completed within 3-5 days. Blood cultures were also prepared for endogenous cases.
PCR could not be performed because PCR analysis for microbial DNA was not available in our institute.

Endophthalmitis treatment protocol
Suspected endophthalmitis was de ned as any case in which the examining physician evaluated the clinical presentation as suggesting an infection and performed a vitreous/aqueous tap, followed by PPV and/or IVI.
At the rst clinic visit, if a patient's retinal re ex was good and their macula, optic disc, and retinal vascular structures could be observed, IVI was applied as the primary treatment. If a patient's retinal re ex was blurred, and their macula, optic disc, and retinal vascular structures could not be observed, PPV was applied as the primary treatment. The clinical course of patients who received IVI was monitored closely for 24 h. If the patients could respond to the treatment, we repeated the IVI after 48-72 h; otherwise, PPV was performed. Kuhn et al. 10 proposed this management scheme as "complete and early vitrectomy".
The patients were prescribed topical forti ed antibiotic drops in addition to IVI or PPV. The drug administered intravitreally was also given as forti ed drops. Topical steroids were applied after the patients responded to antibiotic therapy (i.e., when anterior chamber reaction and vitreous condensation decreased). Systemic and intravitreal steroids were not used. Topical cyclopentolate was used for a mydriatic effect. In both groups 1 and 2, the systemic antibiotic treatments of all patients were started and maintained by consultation with the department of infectious diseases. The drugs given intravitreally were also administered systemically. The patients were discharged after the anterior and posterior segment ndings completely regressed. Therefore, the recovery time of patients was de ned as the hospitalization time.
As a surgical procedure, 23-gauge PPV was performed following brin/exudate removal from the anterior chamber, posterior hyaloid removal, and an injection of 1000 centistoke silicone oil at the end of the surgery. Lensectomy was also applied to three cases as required. At the end of the vitrectomy surgery, intravitreal antibiotics and antifungal injections were applied.
Endophthalmitis was considered culture positive if a positive Gram stain and/or positive growth on culture plates was reported by the institutional microbiology laboratory. Endophthalmitis was considered culture negative when both the Gram stain and culture plates were negative.
All the patients had DM and/or HT as systemic comorbidities. Patients with DM and/or HT were placed under comorbidity group 1. Patients with chronic distant organ in ammation in addition to DM and/or HT were placed under comorbidity group 2.
The culture results, comorbidities, hospitalization times, treatment management, and vision gains of the groups were investigated. The effect of culture results and comorbidities on hospitalization time, treatment management, and vision gain was analyzed.

Statistical Analyses
Statistical Package for the Social Sciences (SPSS) version 22.0 sofware program was used for statistical analysis. Descriptive statistics are presented as minimum, maximum and mean ± standard deviation. A Shapiro-Wilk test was used to normal distribution compatibility of parameters. The parameters did not match to the normal distribution. Therefore, non-parametric tests were used for statistical analysis. A Wilcoxon test was used to in the comparison of initial and nal VA in groups. A Mann-Whitney U test was used to compare comorbidity groups 1 and 2 in terms of hospitalization time, number of IVIs, and visual gain. Chi-square test was used to compare comorbidity groups 1 and 2 in terms of primary therapy, PPV requirement, RePPV requirement. Given a 95% con dence interval, p-values lower than 0.05 indicated a statistically signi cant difference.
Intraoperative complications in surgical procedures were documented in 1/25 eyes (nucleus drop in cataract surgery). There were no intraoperative complications in other surgical procedures.

2) Comorbidity
There were 13 patients in comorbidity group 1 and 12 patients in comorbidity group 2. Comorbidity group 2 showed chronic obstructive pulmonary disease in 4 patients, coronary artery disease in 2 patients, chronic kidney failure in 2 patients, thrombophlebitis in 1 patients, cholecystitis in 1 patients, chronic pyelonephritis in 1 patients, and asthma in 1 patients.
Whatever drug was given intravitreally, the same drug was given systemically as an adjuvant in consultation with the infectious diseases department.
One eye underwent rePPV (4%). This eye was a case where the posterior hyaloid could not be removed, which resulted in an incomplete vitrectomy in the rst surgery. A rePPV was planned for this eye. The posterior hyaloid was removed, and a complete vitrectomy was performed later. P. aeruginosa was isolated in this eye.
The reason for the low VA of 8/9 eyes whose nal VA was less than 0.1 (decimal VA) was due to primary diseases in addition to endophthalmitis. The other eye had severe P. aeruginosa endophthalmitis. Phthisis bulbi occurred due to late complete vitrectomy in this eye. Primary diseases of 8 eyes are shown in Table 1.
No retinal detachment and 1 Phthisis bulbi (the patient with P. aeruginosa endophthalmitis because of late complete vitrectomy) was observed. Evisceration and enucleation were not required in any eye. Table 1. Primary diseases of 8/9 eyes whose nal visual acuity was less than 0.1 in exogenous endophthalmitis 6) The effect of culture results The effect of culture results on hospitalization time, treatment management, and vision gain could not be evaluated statistically because the number of culture + cases was low in exogenous endophthalmitis.

7) Comparison of comorbidity groups 1 and 2
No statistically signi cant difference was found between comorbidity groups 1 and 2 in terms of primary therapy, PPV requirement, RePPV requirement, number of IVIs, and vision gain in exogenous endophthalmitis. However, hospitalization time in comorbidity group 2 was 43% longer than in comorbidity group 1 (p = 0.010). Also, the mean number of IVIs in the comorbidity group 2 was 28% more than in comorbidity group 1. The comparison of comorbidity groups 1 and 2 in exogenous endophthalmitis is shown in detail in Table 2.

2) Comorbidity
There were 3 patients in comorbidity group 1 and 12 patients in comorbidity group 2. Comorbidity group 2 showed chronic kidney failure in 4 patients, chronic pyelonephritis in 3 patients, rectal cancer in 1 patient, hypothyroidism in 1 patient, arrhythmia in 1 patient, prosthetic heart valve in 1 patient, and coronary artery disease in 1 patient.

4) Treatment
Whatever drug was given intravitreally, the same drug was given systemically as an adjuvant in consultation with the infectious diseases department.
There were not retinal detachment and phthisis. Evisceration and enucleation were not required in any eye.

6) The effect of culture results
The effect of culture results on hospitalization time, treatment management and vision gain could not be evaluated statistically because the number of culture + cases was low in endogenous endophthalmitis.

7) Comparison of comorbidity groups 1 and 2
No statistically signi cant difference was found between the comorbidity groups 1 and 2 in terms of hospitalization time, primary therapy, PPV requirement, RePPV requirement, number of IVIs, and vision gain in endogenous endophthalmitis. However, hospitalization time in comorbidity group 2 was 53% longer than comorbidity group 1. Also mean number of IVIs in comorbidity group 2 was 86% more than comorbidity group 1. The comparison of comorbidity groups 1 and 2 in endogenous endophthalmitis is shown in detail in Table 3. In the exogenous endophthalmitis group; The mean hospitalization time was 7.3 ± 2.5 (4-12) days in the comorbidity group 1 (n = 13). The mean hospitalization time was 10.5 ± 2.8 (6-14) days in the comorbidity group 2 (n = 12). There was a statistically signi cant difference between the two groups (p = 0.010). Hospitalization time in the comorbidity group 2 was 43% longer than the comorbidity group 1.
In the endogenous endophthalmitis group; The mean hospitalization time was 14.3 ± 10.1 (8-26) days in the comorbidity group 1 (n = 3). The mean hospitalization time was 22.0 ± 6.7 (12-30) days in the comorbidity group 2 (n = 12). There was not a statistically signi cant difference between the two groups (p = 0.180). However, hospitalization time in the comorbidity group 2 was 53% longer than the comorbidity group 1.
Total group; The mean hospitalization time was 8.6 ± 5.1 (4-26) days in the comorbidity group 1 (n = 16). The mean hospitalization time was 16.2 ± 7.7 (6-30) days in the comorbidity group 2 (n = 24). There was a statistically signi cant difference between the two groups (p = 0.000). Hospitalization time in the comorbidity group 2 was 88% longer than the comorbidity group 1. The mean hospitalization times of the groups are shown in Table 4.

Discussion
In this study, we evaluated the clinical outcomes of endophthalmitis. It also evaluated the effect of changes of comorbidities on hospitalization, treatment management, and vision gain.

Group 1 (exogenous endophthalmitis)
The culture positivity rate in large series was reported to be 45%-75%, [11][12][13][14][15][16] and this rate was increased by polymerase chain reaction (PCR) in exogenous endophthalmitis. 14 4/25 eyes (16%) had culture positivity in this study. There are several common causes of culture negativity. Identi cation of the causal pathogen from culture media is limited due to the early administration of broad-spectrum or prophylactic antimicrobial drugs, as well as organisms that are fastidious or slow growing. Also, the differences in methods of taking culture samples may affect the culture results. In this study, aqueous and vitreous samples were obtained as a culture sample, as in the Endophthalmitis Vitrectomy Study (EVS). 17 The samples were placed on culture mediums (Fluid thioglycolate medium, Blood agar, Chocolate agar, Eosin methylene blue agar, and Sabouraud dextrose agar) in the operating room. Despite this, we found that our culture positive rate was low compared with those described in the literature. [11][12][13][14][15][16] In a multicenter study conducted by the European Vitreo-Retinal Society (EVRS) endophthalmitis study group in 2019, in postoperative endophthalmitis, 45% and 54% of the culture positivity were in two different groups, 26.1% of the eyes required two IVIs, and 12.2% of the eyes required more than two IVIs. 16 In this study, the culture was 16% positive, 52% of the eyes required two IVIs, and 20% of the eyes required more than two IVIs. In this study, more repeat intravitreal injections were necessary than in the multicenter study conducted by the EVRS endophthalmitis study group. We believe this was due to the low culture positivity rate and the lack of detection of the most sensitive antibiotics.
In the group undergoing vitrectomy in the EVS and the Complete and Early Vitrectomy for Endophthalmitis (CEVE) study, rePPV rate was 0%. 10,17 In the this study, the rePPV rate was 4% (one eye). The eye that applied rePPV is a case where posterior hyaloid could not be removed and incomplete vitrectomy in the rst surgery. P. aeruginosa were isolated in this eye; rePPV was planned; the posterior hyaloid was removed; and a late complete vitrectomy was performed. Unfortunately, phthisis bulbi occurred due to late complete vitrectomy in this eye.
In the literature, different rates related to visual outcomes in endophthalmitis have been reported. In the EVS, 17 82% of eyes had a nal VA of 0.1 or better, whereas in this study, 64% of eyes had a nal VA of 0.1 or better. In this study, low vision was attributable to underlying primary diseases in addition to endophthalmitis, as indicated in Table 1, in eyes with a nal VA less than 0.1. Only in one eye, phthisis bulbi occurred due to late complete vitrectomy, and the nal VA remained less than 0.1.
The hospitalization time in the comorbidity group 2 was 43% longer than the comorbidity group 1, and the mean number of IVIs in the comorbidity group 2 was 28% more than the comorbidity group 1. Since the culture results and antibiotic sensitivity were unavailable, it was determined that the comorbidity group 2 with underlying diseases, in addition to HT and/or DM, needed a signi cantly longer time to recover.

Group 2 (endogenous endophthalmitis)
According to the endogenous endophthalmitis literature review, in a study by Connell et al, 2 64.1% of eyes had a positive ocular culture; 46.8% of eyes had a nal VA of 0.1 or better; and 7.8% of eyes were enucleated. In two studies by Jackson et al, 18,19 58% of eyes had a positive ocular culture; 56% of patients had a positive blood culture; 26%-49% of eyes had a nal VA of 0.1 or better; and 19%-25% of the eyes were enucleated and eviscerated. In a study by Ratha et al, 20 58.6 % of eyes had a positive ocular culture; 29.5% of eyes had a nal VA of 0.1 or better; and 19.7 % of the eyes were eviscerated. In this study, in the endogenous endophthalmitis group, 20% of eyes had a positive ocular culture; 0% of patients had a positive blood culture; 53.3% of eyes had a nal visual acuity of 0.1 or better; and 0% of eyes had evisceration and enucleation. Compared to the results of these studies, our results are superior, although our culture isolation rate is quite low. In a study by Jackson et al, 18 PPV+ intravitreal +systemic treatment was applied to 20% of eyes; 9% of eyes undergoing PPV required evisceration and enucleation; and 26% of eyes that did not undergo PPV required evisceration and enucleation. In a study by Connell et al, 2 43.75% of eyes applied PPV; enucleation was not performed in the PPV group; and 16.1% of the eyes were enucleated in the nonPPV group. In a study by Ratha et al, 20 62.3% of eyes underwent PPV, and 13.1% of eyes required rePPV.
In this study, PPV was applied to 13/15 eyes (86.7%), and rePPV was applied to 3/15 eyes (20%). All patients received intravitreal and systemic treatment. We observed that PPV combined with intravitreal and systemic treatment in endogenous endophthalmitis increases vision gain and decreases evisceration/enucleation. We found that our culture positive rate (20%) increased our rePPV rate (20%) in endogenous endophthalmitis cases. If the culture results were more bene cial, the need for rePPV could be reduced with more sensitive antibiotherapy-antifungal therapy.
The mean hospitalization time comprised 14.3 ± 10.1 (8-26) days for the comorbidity group 1 (n = 3). The mean hospitalization time was 22.0 ± 6.7 (12-30) days for the comorbidity group 2 (n = 12). There was no statistically signi cant difference between the two groups due to a low number of patients. However, the hospitalization time in the comorbidity group 2 was 53% longer than the comorbidity group 1, and the mean number of IVIs in the comorbidity group 2 was 86% more than the comorbidity group 1.
Since the culture results and antibiotic sensitivity were unavailable, it was determined that the comorbidity group 2 with underlying diseases, in addition to HT and/or DM, needed a signi cantly longer time to recover.
This study showed that clinical conclusions of endophthalmitis indicate that early complete PPV decreases endophthalmitis complications in exogenous endophthalmitis. PPV combined with intravitreal and systemic treatment in endogenous endophthalmitis increases vision gain and decreases evisceration/enucleation. Because microorganisms cannot be isolated in culture, repeat IVI was found at a higher rate than that in the literature on exogenous endophthalmitis, and rePPV was found at a higher rate than that in the literature on endogenous endophthalmitis. In these studies in the literature, the effect of comorbidity has not been investigated, and information about the complete recovery times of the cases is lacking.
According to the endophthalmitis literature review, in studies investigating comorbidity in patients with endophthalmitis, the most common comorbidities were reported as DM, HT and leukemia / lymphoma in endogenous endophthalmitis. [21][22][23] In the study of Weng et al, 24 DM and HT have been reported as the most common comorbidities in endophthalmitis patients. Studies of the relationship between the comorbidities of patients with endophthalmitis and hospitalization time are lacking. In the study of Weng et al. 24 it has been reported patients with renal disease, septicemia, pneumonia, and tumor had a higher mortality rate. In addition, it has been reported hospitalization time in the mortality group was higher than in the survived group. However, the relationship between comorbidities and hospitalization time has not been reported.
Duric et al. 5 reported systemic comorbidity increases psychological and physiological stress in patients, and due to stress, cytokine release such as interleukin (IL)-1beta, IL-6, and tumor necrosis factor (TNF) alpha increases in ammation. In addition, It has been stated that oxidative stress and in ammation trigger each other. 25,26 Kawashima et al. 6 presented systemic comorbidity increases oxidative stress, and due to oxidative stress, ocular diseases associated with oxidative stress and in ammation such as dry eye disease increases. Fang et al. 7 presented systemic comorbidity is associated with prolonged the clinical course of critical patients and prolonged hospitalization time in patients with COVİD-19. In addition, Guan et al. 27 reported a greater number of comorbidities is correlated with poorer clinical outcomes due to prolonged in ammation in patients with COVİD-19. Our study showed that patients with chronic distant organ in ammation in addition to DM and/or HT encountered increased hospitalization times due to prolonged in ammation. The hospitalization time is the recovery time and was higher in the comorbidity group 2 compared to the comorbidity group 1 in the exogenous group, endogenous group and total group. Studies of the relationship between the comorbidities of patients with endophthalmitis and hospitalization time are lacking. We assert that this study will be the rst investigation to address this subject. The research ndings suggest that patients with chronic distant organ in ammation in addition to DM and/or HT encountered increased hospitalization times.

Conclusions
Because our culture and antibiogram ndings were low, we had to increase surgical procedures and repeat IVI. This study showed that patients with chronic distant organ in ammation in addition to DM and/or HT encountered increased hospitalization times and IVI requirements.

Declarations
Acknowledgments Systemic antibiotic treatment was initiated and continued with consultation with the department of infectious diseases. Thank you for your help to the department of infectious diseases.
Compliance with Ethical Standards: Con ict of Interest: Author AF declares that he has no con ict of interest. Author HG declares that she has no con ict of interest. Author TO declares that she has no con ict of interest.
Financial Disclosure: Author AF declares that he has no nancial disclosure. Author HG declares that she has no nancial disclosure. Author TO declares that she has no nancial disclosure.
This study conforms with the Helsinki Declaration principles. This study was approved by the local ethics committee. Tables   Table 1. Primary diseases of 8/9 eyes whose nal visual acuity was less than 0.