Schirmer strip and conjunctival swab for viral detection on the ocular surface of adults: a scoping review

Purpose Schirmer Strips and Conjunctival swabs are used in ophthalmology for tears and uids collection. During the COVID-19 pandemic, one of the biggest challenges is accurate diagnosis, and it is known that, in some cases, ocular manifestations are one of the rst symptoms. In this context, this study has the objective of raising scientic evidence that highlights the use of Schirmer strips and conjunctival swabs as a method of sample collection for viral analysis to support future research on this theme. Methods A literature search was conducted in the PubMed, Web of Science and BVS databases, following the Scoping Review protocol dened by Joanna Briggs Institute (JBI) after the guiding question “Is it possible to detect viruses on the ocular surface with Schirmer Test and/or conjunctival swab?”. Results A total of 418 studies were identied, and after discerning analysis, 36 English written studies were selected. Three researchers analyzed studies after virus research, collection methods, and sample analysis. Publications were mainly on adenovirus, herpes simplex virus and SARS-CoV-2, and there is also evidence of ocular detection of more viruses types. Conjuntival swab analyzed through Polymerase Chain Reaction (PCR) or Reverse Transcriptase-PCR (RT-PCR) were the most used methods. Conclusions Studies have generally been conducted to understand viral infection, to develop accurate diagnostic methods and to follow the patients’ response to treatment. Most studies were performed with a small number of patients and lacked clear denitions of collection time and viral persistence since the onset of diseases. Viruses can be detected on the ocular surface through the analysis of Schirmer strips and conjunctival swabs. However, additional studies with larger populations and time permanence are necessary to develop more assertive conclusions on the theme.


Introduction
Viruses are exclusively intracellular parasites and are also the smallest known infectious agents. Its mechanisms of disease are still not completely clear; however, there are direct factors that contribute to viral tropism: viral receptors in the host cell, speci c cell line, and physical barriers that enable and/or inhibit infections. Once inside the cell, the virus may damage or destroy it through direct cytophatic effects, host antiviral immune responses, and/or transformations of the infected cells (1) .
It is known that the eye is a site for viral infections that might appear in the intra-or extraocular space without visible systemic reverberation and affect multiple structures with variable manifestations (2,3) . Red eye, pain and blurred vision are one of the rst clinical signs and symptoms of viral ocular impairment. Conjunctivitis is the topmost ocular infection in primary health care, given that 65-90% of cases are caused by adenovirus (AdV) and occasionally herpes simplex virus (HSV) or varicella zoster virus (VZV). Keratitis caused by HSV-1 is also commonly seen, and it is still the main cause of blindness succeeding infection in developed countries, presenting approximately 40 thousand new cases of visual impairment every year worldwide (4) . Viruses that are frequently associated with ocular or systemic complications are Epstein Barr virus (EBV), Measles morbilivirus and Paramyxovirus (5) . Viral diagnosis is made after clinical signs and symptoms and laboratory results that support medical hypotheses. There are a large variety of laboratory tests, but speci city and sensitivity change from one microorganism to another. The doctor in charge, based on clinical information and previous experiences, should decide between available options considering the patient's singularities. Cell culture and analysis of genetic material with samples collected from blood, mucosa or secretions are the main methods. In the eye, the most commonly used collection method is conjunctival swabs, but Schirmer strips have also shown good results.

Current scenario
The World Health Organization (WHO) declared COVID-19 a pandemic threat, because of that, the academic community has concentrated all efforts to solve the pandemic and give people's normal lives back. The Chinese Hero and ophthalmologist Li Wenliang, MD, rst reported the possibility of a novel catastrophic virus, SARS-CoV-2, and now, it is already known that, in some cases, ocular manifestations are one of the rst symptoms, and consequently, the eye may contribute to the understanding of COVID-19 pathophysiology (6,7) . Since then, several studies on ocular manifestations of SARS-CoV-2 have been conducted, but virus collection methods and associations with the ocular surface are not clearly stated.

Schirmer test
The idea of collecting tears as a clinical test was rst introduced by the German ophthalmologist Köster in 1900. The test consisted of the placement of lter paper on all extensions of the conjunctival sac while the nasal mucosa was stimulated to produce tearing caused by nasal irritation. The objective was to exhaust tear production to evaluate the function of lacrimal glands. Therefore, this test could take up to 90 minutes, becoming exhaustive and unviable for daily medical practice (8) .
In 1903, Otto Schirmer, also a German ophthalmologist, shortened the size of the paper strips and quanti ed tear production for 5 minutes by three methods, each one analyzed a distinct tearing stimulation pathways: ocular and palpebral mucosa, nasal mucosa, and the retina (8,9) . Since then, several modi cations have been proposed on the Schirmer test; however, this test remains important in the quanti cation and standardization of tear volume. Currently, it is made with a lter paper strip 60 mm long and 5 mm wide, which is inserted in the temporal side of the conjunctival sac. The patient's eyes are closed, the strips are removed after 5 minutes, and the wet part is measured. Normal values are considered to be results greater than 15 mm, but they can vary mainly according to medication use, age and the presence of chronic diseases (10,11) .

Conjunctival swab
Conjunctival swabs are the most commonly used methods for microbiological analysis because they allow the collection of cells and materials that are spread in the conjunctival sac instead of tears exclusively. The only exception are the in-o ce rapid antigen tests for AdV, since they are faster and they also present a good sensitivity and speci city, 89% and 94%, respectively (12) . The method of collection is also very fast and simple: a swab with a cotton tip is gently passed in rotational movements on the conjunctival sac (13) . Topic anesthesia can be used to make the procedure more confortable, since there is no signi cant difference in the nal results when samples are analyzed by polymerase chain reaction (PCR) methods (14) . Proxymetacaine 0.5% is recommended for ocular surface anesthesia since it shows fewer bactericidal effects among the commercially available eye drops (15) . It is also recommended to use sterile swabs, since calcium-containing swabs can inhibit polymerase activity (16) .
In this context, this study has the objective of raising scienti c evidence that highlights the use of on Scoping Reviews (17) . All searches and publication access were completed in June 2020, and had no time limit.
The guiding question "Is it possible to detect virus on the ocular surface with Schirmer Test and/or conjunctival swab?" was de ned for the selection and the search of the studies. This was built through the PCC strategy, which consists of a mnemonic for the words Population, Concept and Context. In this way, "P" was de ned as adult patients (older than 18 years old), "C" as Schirmer Test and conjunctival swab, and the last "C" as all viruses.
The search was performed using the databases PubMed, Web of Science, and Blibioteca Virtual em Saúde (BVS). Included articles were only those written in English, published in indexed sources, and with quantitative or qualitative approaches, primary studies and reviews.
Insightful reading of the title, abstract and key words was performed to select the articles according to previously established inclusion and exclusion criteria. When the title, abstract and key words were not su cient, the full text was also analyzed. All articles were called studies, enumerated in chronological order, and evaluated by three different researchers. Recommendations by JBI were adapted for the study singularities and used for data extraction. This article followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guideline and checklist developed under EQUADOR (Enhancing the QUAlity and Transparency Of health Research) network guidance (18) .

Results
Following the database search, 418 potential studies were identi ed. After reading the title, abstract and key words, 79 studies were selected, and 27 were excluded because they were also found in different databases. The full texts of the 52 remaining articles were read, and 16 were excluded for not answering the guiding question. Using the described methodology, a literature search found 36 articles that met all criteria. This process is shown in Fig. 1.
The three researchers analyzed all studies after virus research, collection methods, and sample analysis.
Most of them were experimental or observational studies developed in the United Kingdom (7/36). Studies on the topic started in 1997, but 55,6% were only published in the decade of 2011-2020.
The preferred collection method was conjunctival swab alone, 29 out of 36 articles, while six studies used Schirmer strips, and only one of them compared both collection methods (26) . One of the studies was a narrative review about SARS-CoV-2 (46) .
The majority of the samples were analyzed by PCR or Reverse Transcriptase-PCR (RT-PCR) with the exception of the studies number 1 (19) and 17 (35) , that used immunochromatography (IC) and enzyme immunoassay (EIA), and immunoblot analysis, respectively.
Details of the results are speci ed in Fig. 3-5.

Discussion
A literature search found that viruses might be identi ed on the ocular surface through the analysis of conjunctival swabs or Schirmer strips, but studies mainly focused on AdV, HSV and SARS-CoV-2, while other viruses were still not deeply investigated.
Studies that focused on the detection of AdV and HSV were linearly developed over the last 20 years and randomly conducted all over the world. In contrast, in 2020, scientists exponentially published studies on SARS-CoV-2 because the virus was rst seen in the last days of 2019; in a few months, the WHO declared its disease as a pandemic, and the world's research efforts were completely directed to overcoming this pandemic. The evolution of published articles on AdV, HSV and SARS-CoV-2 is represented in Fig. 2.
In contrast to SARS-CoV (54) , SARS-CoV-2 RNA was found by RT-PCR in conjunctival swabs (40)(41)(42)(43)(44)(45)(46)(47)(48) , although samples were collected from a few patients, and only a low and varying percentage of them presented positive results and/or ocular symptoms (46) . Most of these studies were developed in China (40,41,43,47) , mainly because the country was the rst epicenter of the disease. In some cases, ocular manifestations are one of the rst symptoms, and because of that, academic societ believes the eye may contribute to the understanding of COVID-19 pathophysiology (6,7) . Recently, new studies proved the presence of in the conjunctiva, limbus, and cornea, with prominent staining in the super cial epithelium surface, key factors for the infection of SARS-CoV-2 in human cells (55) . Those results collaborate for the necessity of ocular protection in preventing the spread of viruses.
Research on viral screening was, most of the time, correlated with external ocular symptoms with the purpose of solving clinical doubts between pathogenic agents and of looking after methods for a fast accurate viral diagnosis. Studies on VZV, CMV and EBV were mostly combined with HSV to determine coinfection and differential diagnosis (32)(33)(34) .
Only one study correlated viral detection on the ocular surface with intraocular symptoms. This study was published in 2000 and aimed to verify the e cacy of intravenous ganciclovir treatment in immunocompromised AIDS patients with CMV retinitis. The results showed a high clinical relevance for con rming and differentiating the diagnoses of CMV retinitis when ophthalmoscopic ndings were associated with PCR methods on conjunctival swab samples (49) . No other studies on the topic were found in this review, probably because the incidence of CMV retinitis in the Acquired Immunode ciency Syndrome (AIDS) population signi cantly decreased with the introduction of effective antiretroviral therapy and early accurate diagnosis (56)(57)(58)(59) .
The most commonly used method for sample collection was conjunctival swabs. This method also collects conjunctival cells, while Schirmer strips allow only the collection of tears, since uids pass to the lter paper because of gravity, viscosity, and capillary ow dynamics -the same physical processes that explain how liquids impregnate porous materials differently (60,61) . Consequently, the samples collected represent two different materials: tears, cells and uids dispersed in the conjunctival sac and tears and substances dissolved in it.
Samples were mainly analyzed by PCR or RT-PCR, depending on the viral genetic material, DNA or RNA, respectively. PCR variations are widely used because they allow the replication and detection of low loads of viral DNA/RNA (62) . The point-of-care test was also compared with PCR effectiveness, sensitivity and speci city where high, but this eld was only explored in AdV (30) , probably because of epidemiological factors related to uncontrolled and fast spread of AdV conjunctivitis (63) .
This review included a large number of articles, 36, from many different countries and time periods, fact that collaborated for a broad comprehension of how the detection of viruses on the ocular surface was worldwide studied in the last decades. The results of this article may contribute for future research by clarifying key concepts that might assist the design of future researches on ocular viruses.
In conclusion, viruses can be detected through the analysis of samples collected by Schirmer strips and conjunctival swabs, and studies were generally conducted to understand viral infection, to develop accurate diagnostic methods and to follow patients' response to treatment. However, this study has some limitations: only three databases were consulted, and new articles on the theme are constantly being published, which might exclude relevant outcomes. Moreover, among the analyzed studies, most were developed on a small number of patients and lacked clear de nitions of collection time and viral persistence since the onset of the diseases. Additional studies with larger populations and time permanence are necessary to develop more assertive conclusions on this topic.
Author's contributions LES, AM and LFML designed the work, acquired the data, contributed to the analysis and interpretation of data, and drafted the paper. JS, TETO and CFS contributed to the analysis and interpretation of data, and revised the paper. All authors approved the nal version.