FormalPara Key Summary Points

Why carry out the study?

Studies find that patients report a certain degree of fear and anxiety regarding intravitreal injection, and concerns related to pain from intravitreal injections are one of the key factors mentioned by patients when asked about therapy.

We systematically review comparative studies on the needle gauge size used for intravitreal anti-vascular endothelial growth factor injection therapy and the pain experience reported by the patients.

What was learned from the study?

Our network meta-analysis found a clear relationship between thinner needle gauge sizes and lower pain experience from intravitreal injections.

However, pain is a complex experience, and studies report many influencing aspects related to intravitreal injections. Nevertheless, we recommend using the thinnest feasible gauge size when delivering intravitreal injection therapy.

Introduction

The introduction of intravitreal anti-vascular endothelial growth factor (anti-VEGF) therapy has dramatically improved the prognosis of exudative maculopathy [1, 2]. In the USA alone, it is estimated that 5.9 million intravitreal injections were performed in 2016 [3]. The number of intravitreal anti-VEGF injections is growing. Jørstad et al. from Norway reported a 100-fold increase in the number of annual intravitreal anti-VEGF injections from 2006 to 2018 [4], and Thinggaard et al. estimate that the number of annual intravitreal anti-VEGF injections will increase by 50% in Denmark within 5 years [5]. Considering these large figures, it makes great sense to focus on optimizing the procedure to create an efficient, safe, and welcoming service that delivers as painless an intravitreal injection as possible.

Studies find that patients report a certain degree of fear and anxiety regarding the intravitreal injection [6,7,8,9], and development of anxiety due to the expectation of pain from the injections [10]. Patients even report sleep disturbance as a result [11], and concerns related to pain from intravitreal injections are one of the key factors mentioned by patients when asked about therapy [12,13,14]. Although most patients remain in therapy despite these challenges, a minority opt out of treatment due to the burden from repeated intravitreal injections [15]. Hence, guidelines for intravitreal injections often focus on approaches to limit the associated experience of pain [16,17,18].

One aspect regarding the pain experience is the gauge size of the needle used for intravitreal anti-VEGF injection therapy [16]. In this systematic review, we evaluated the pain experience in relation to gauge size in intravitreal anti-VEGF therapy to provide an overview of the evidence in this field. Further, we used network meta-analysis as a method to compare pain reports across studies and ranked gauge sizes according to pain severity.

Methods

Study Design

This study was a systematic review and network meta-analysis designed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [19]. Recommendations of the Cochrane Handbook were followed for practical and analytical guidance [20]. The study protocol was registered in the PROSPERO database (ID: CRD42023474162). This study is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Study Eligibility Criteria and Outcome of Interest

We employed the following eligibility criteria when considering studies for this review:

Population

Patients undergoing intravitreal anti-VEGF therapy with intravitreal injections were included. We did not enforce any criteria on the decision for intravitreal anti-VEGF therapy. No restrictions were made on patient characteristics (e.g., age, sex, race) or any eye characteristics (e.g., phakic/pseudophakic, underlying disease, visual acuity). We did not enforce any restrictions on the anti-VEGF drug employed for such a study; however, intravitreal anti-VEGF therapy considered outside the definition of routine intravitreal anti-VEGF therapy was not eligible for this study (e.g., port delivery devices, experimental delivery systems, implant-based systems). Intravitreal therapy without any anti-VEGF or a combination of anti-VEGF with any other drug category (e.g., antibiotics or corticosteroids) was not considered eligible for this study. We did not enforce any restrictions on the protocol for topical anesthesia but noted such details for the qualitative review.

Intervention and Comparison

Intravitreal injection using any gauge size. Comparison groups were defined as other gauge sizes. Studies without any data on a comparison group were not considered eligible.

Outcomes

Our outcome of interest was the pain experience. We did not apply restrictions on the measure used to report on pain experience. We required that pain was reported by the patient (i.e., not a bystander or the injection provider) and that pain was measured immediately after the injection procedure.

Study Type

We only considered comparative controlled trials. We did not enforce any restrictions based on blinding strategy, randomization, or any other methodological characteristics. We only considered peer-reviewed studies reported in the English language. We did not consider conference abstracts or non-peer-reviewed studies. We did not restrict eligibility based on geography or journal.

Information Source, Search Strategy, and Study Selection

We searched 12 literature databases on 14 October 2023: PubMed, Embase, the Cochrane Central Register of Controlled Trials, Web of Science Core Collection, BIOSIS Previews, Current Contents Connect, Data Citation Index, Derwent Innovations Index, KCI–Korean Journal Database, Preprint Citation Index, ProQuest™ Dissertations & Theses Citation Index, and SciELO Citation Index. All searches were conducted by one trained investigator (author Yousif Subhi) using search phrases specifically tailored to the individual databases (details are available in Supplementary Material).

Records extracted from the literature search were imported to EndNote X9.3.1. for Mac (Clarivate Analytics, Philadelphia, PA, USA). All records were screened by one author (Yousif Subhi) to remove duplicates and those obviously irrelevant. The remaining records were retrieved for full-text evaluation of eligibility by two authors working independently (Emilie T.S. Butler and Jakob Bjerager). Reference lists were checked for any eligible records not identified by the literature search. The authors then discussed study selection and eligibility with a third author (Yousif Subhi), who provided a final decision if consensus could not be reached.

Data Items, Data Collection, and Risk of Bias within Studies

From each eligible study, we extracted data on study design, study characteristics, study population, intervention/comparisons, and outcomes. Risk of bias within studies was evaluated using the Cochrane risk-of-bias tool version 2 [21]. Two authors worked independently to extract data from studies (Emilie T.S. Butler and Nathalie S. Eriksen) and to evaluate the risk of bias within studies (Emilie T.S. Butler and Jakob Bjerager). The authors then discussed the data extraction and risk-of-bias evaluation with a third author (Yousif Subhi), who provided a final decision if consensus could not be reached.

Data Synthesis and Risk of Bias across Studies

All studies were reviewed qualitatively in text and in tables. For the quantitative analyses, i.e., the network meta-analyses, we constructed network plots to illustrate the network of direct comparisons and to confirm the existence of a complete network necessary for the analyses. We then used the generalized pairwise modeling approach, which is based on the repeated application of adjusted indirect comparisons [22]. The generalized pairwise modeling approach has been shown to deliver robust results and is comparable to the Bayesian approach and the multivariate modeling approach [22]. All meta-analyses were conducted using MetaXL 5.3 (EpiGear International, Sunrise Beach, QLD, Australia) for Microsoft Excel (Microsoft, Redmond, WA, USA), which uses the generalized pairwise modeling framework. All outcomes of pain were transformed into standardized effect sizes using Cohen’s d, which is calculated using the ratio of the mean difference between the groups and the standard deviation for the data. The random-effects model was employed to account for potential heterogeneity across studies. Summary estimates were given using 95% confidence intervals (95% CI). Statistical significance was defined as P values less than 0.05.

Results

Study Selection

The literature search identified 1809 records. One record known to us was added to the pool of references. From a total of 1810 records, we removed 396 duplicates and 1402 obviously irrelevant records. The remaining 12 records were retrieved in full text for further evaluation of eligibility. No additional records were identified by reviewing reference lists. We found nine studies to be eligible for both the qualitative and the quantitative review (Fig. 1).

Fig. 1
figure 1

PRISMA flow diagram of study selection

Full size image

Characteristics of the Eligible Studies

Studies collectively summarized data on a total of 998 patients and 1004 eyes [23,24,25,26,27,28,29,30,31]. Data were reported from participants in Europe (4 studies), Asia (2 studies), North America (2 studies), and South America (1 study). Where reported, eyes were treated for a variety of retinal diseases, most commonly age-related macular degeneration, diabetic macular edema, and retinal vein occlusion. Treatments were performed using bevacizumab [26,27,28,29, 31], ranibizumab [23,24,25,26, 30], or aflibercept [23, 25, 30]. Pain experience was evaluated using a 10-point numeric rating scale, a 10-point visual analogue scale (VAS), a 5-point Likert scale, and the Wong-Baker FACES scale. These aspects of pain were measured immediately after the injection in all studies, and one study also measured pain later during the day and the following day [25]. The intravitreal injection protocols differed across studies in terms of pre-injection preparation protocol and injection site. Injections were given by physicians or surgeons. Further study characteristics and details of injection protocols are summarized in Table 1.

Table 1 Study characteristics
Full size table

Needle sizes in studies included 26-gauge (1 study), 27-gauge (4 studies), 29-gauge (1 study), 30-gauge (all 9 studies), 32-gauge (1 study), 33-gauge (3 studies), and 34-gauge (1 study). Seven studies reported patient demographics, and these studies reported a mean age of 60–78 years and a balanced distribution of male and female sex. Further details of the study groups are summarized in Table 2.

Table 2 Characteristics of study groups in reviewed studies
Full size table

Results of Individual Studies

Güler et al. [26], Haas et al. [27], and Lourerio et al. [28] compared the pain experience of treatment with 27-gauge and 30-gauge needles. Güler et al. found that the thinner 30-gauge needle was associated with lower pain scores, but that the difference in pain experience between the gauge sizes was age-dependent, as it was only observed in patients aged < 65 years [26]. Haas et al. found no difference in the pain experience between the two gauge size groups, but reported that the pain experience was positively correlated with higher age, female sex, and number of previous injections [27]. Lourerio et al. found no difference in the pain experience between the two gauge size groups and found no correlation between the pain experience and patient demographics or previous number of injections [28].

Rodrigues et al. compared six types of needles among four needle calibers: 26-, 27-, 29-, and 30-gauge needles [29]. This study found that patients treated with 26-gauge or 27-gauge needles experienced significantly more pain than those treated with 29-gauge or 30-gauge needles [29]. No correlation was found between pain experience and needle length, patient demographics, injection technique, or the temporary intraocular pressure increase after the injection [29].

Five studies compared the 30-gauge needle with thinner needle sizes [23,24,25, 30, 31]. Aderman et al. compared the pain experience of treatment with 30-gauge and 33-gauge needles [23]. Patients received bilateral injections on the same day, one injection with each gauge size, and the order of the gauge size was randomized [23]. This study showed that significantly more patients preferred the thinner 33-gauge needle, but the difference in the pain score between the two gauge sizes was not statistically significant [23]. Alshahrani et al. randomized patients to treatment using either 30-gauge or 32-gauge needles [24], and found a lower pain score in patients treated with the thinner 32-gauge needle [24]. Both Aderman et al. and Alshahrani et al. explored the relationship between age, sex, and diagnosis and the pain experience, and neither study found any significant correlation [23, 24]. Eaton et al. compared the pain experience of treatment with 30-gauge and 33-gauge needles [25]. The patients were asked about their pain level during the injection, immediately after the injection, later the same day, and the next day—no significant differences were found between the two groups [25]. However, this study reported that the topical anesthesia strategy, and especially the use of subconjunctival anesthesia, led to lower pain scores [25]. Sasajima et al. randomized eyes into either 30-gauge or 34-gauge needles and found that patients reported lower pain on pain scores when the 34-gauge needle was used [30]. Van Asten et al. recruited patients with at least two consecutively planned injections and performed one of the injections with a 30-gauge needle and the other with a 33-gauge needle and randomized the order of the gauge sizes used [31]. This study found no statistically significant difference, and no change in the pain experience from one injection to the next [31]. A positive correlation was found between the pain experience and the following: experience of distress prior to the injection, negative view of the preparations for the injection, negative experience with prior injections, expected negative treatment outcomes, and lower belief that the injections would improve their health [31].

Risk of Bias within Studies

Risk-of-bias evaluation was based on assignment to intervention (i.e., the intention-to-treat effect) and made on five domains (randomization process, deviations from intended interventions, missing outcome data, measurement of outcome, and selection of the reported results) which together contributed to an overall risk-of-bias evaluation. The randomization process found a high risk of bias in two studies which did not randomize participants and some concerns in the remaining studies due to the lack of details of the randomization process. Some concerns were raised across all studies in the domain of deviations from intended interventions mainly because individuals delivering the interventions were aware of the participants’ assigned intervention during the study. In the remaining domains, no concerns were found. The risk-of-bias evaluations within studies are summarized in Fig. 2.

Fig. 2
figure 2

Risk of bias of individual studies

Full size image

Network Meta-Analyses

The network plot illustrates a complete network (Fig. 3). We used the largest gauge needle size in the network (26-gauge needle) as the reference group and compared the other gauge needle groups through direct and indirect comparisons within the network (Fig. 3). Pain experience from intravitreal anti-VEGF injections was ranked as follows (from the most to the least painful) and according to the difference in effect size (Cohen’s d) as compared with the 26-gauge needle reference group (Fig. 3):

  • 27-gauge needle (d: −0.4; 95% CI −1.9 to 1.2; P = 0.66)

  • 29-gauge needle (d: −2.7; 95% CI −3.2 to −2.3; P < 0.00001)

  • 30-gauge needle (d: −3.8; 95% CI −5.3 to −2.4; P < 0.00001)

  • 32-gauge needle (d: −4.3; 95% CI −5.2 to −3.5; P < 0.00001)

  • 33-gauge needle (d: −4.5; 95% CI −5.4 to −3.7; P < 0.00001)

  • 34-gauge needle (d: −5.3; 95% CI −6.2 to −4.4; P < 0.00001)

Fig. 3
figure 3

Network meta-analysis of the effect of needle gauge size on pain after intravitreal anti-vascular endothelial growth factor injection. Left: Network plot illustrating comparisons in the meta-analysis. The size of the green dots indicates the number of studies including that specific needle gauge (G) size group. The thickness of the lines between the green dots corresponds to the number of direct comparisons between the needle gauge size groups within the network. Right: Summary estimates (dots and whiskers) for pain experience in each needle gauge size group are provided as standardized effect size (Cohen’s d) relative to that of the 26G group (reference, indicated by the dotted line). A group is not statistically different from the reference if the confidence interval crosses the dotted line

Full size image

Detailed summaries of each individually calculated direct and indirect estimate are available in Supplementary Material.

Discussion

This systematic review and network meta-analysis of needle gauge sizes and pain experience found a clear relationship between thinner needle gauge sizes and lower pain experience from intravitreal injections. According to the interpretation of Cohen’s d, the rule of thumb is that effect sizes are small at d = 0.2, moderate at d = 0.5, and large at d = 0.8 [32]. In that regard, we observe very large differences between the thicker 26-gauge and 27-gauge needles, and the thinner sizes of 30-gauge (d = −3.8), 32-gauge (d = −4.3), 33-gauge (d = −4.5), and 34-gauge (d = −5.3). Considering that intravitreal injections with anti-VEGF often require many treatments over years for patients [33,34,35], improving the patient experience in relation to the treatments is of great importance.

Our findings in this study follow the notions of published guidelines and expert consensus papers [16,17,18, 36,37,38,39,40]. The Euretina Expert Consensus Recommendations state that 30-gauge or thinner needles are recommended, but that larger needles can be employed when necessary [16]. The Royal College of Ophthalmologists state that 30-gauge needles are recommended [36]. The Royal Australian and New Zealand College of Ophthalmologists recommend a 27-gauge needle for triamcinolone and a 30-gauge or thinner for other preparations [37]. The Intravitreal Injection Task Force Committee of the American Society of Retina Specialists recommends a gauge size of 30, 31, 32, or 33 [17]. The American Academy of Ophthalmology notes that 30-gauge needles are commonly used, and thinner needle sizes may decrease patient discomfort [38]. One expert update published in Retina Specialist recommends 30-gauge or thinner [39], and another guideline published in Review of Ophthalmology does not comment on gauge size [40]. One international practical approach paper recommends the use of 30-gauge needles [18].

Studies of pain show that the brain can generate the experience of pain even in the absence of input from peripheral nociceptors [41]. Pain experience is therefore a complex topic in which sensory input from peripheral nociceptors, for example, are processed together with affective and cognitive brain activity [41]. The complexity of the pain experience and the multifactorial influence on pain is also underscored by the definition of pain according to the International Association for the Study of Pain: “an unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage” [42]. In this regard, one would expect measurements of pain across studies to be associated with various factors not related to the injection procedure. Indeed, van Asten et al. explored these aspects in detail [31]. Greater pain experience was reported among patients who had an experience of distress prior to the injection, a negative view of the preparations for the injection, negative experience with prior injections, expected negative treatment outcomes, and lower belief that the injections would improve their health [31]. Thus, addressing pain related to intravitreal injections is a complex endeavor that also requires a holistic focus on the intravitreal injection service and the approach to the patients and addressing their fears, experiences, and expectations.

Limitations should be acknowledged when interpreting the results of this study. First, we only compared gauge sizes and disregarded analytically other aspects of preparation for the intravitreal injection. This limitation still allows comparisons throughout the network, as we are measuring differences between the groups that are randomized/compared according to the gauge size, i.e., the remaining aspects of preparations should remain similar between the groups within study. Still, however, it is likely that a topical anesthesia protocol applied in one study may differ significantly in anesthetic efficacy from a protocol applied in another study, which then may influence the difference between individual groups within studies that consequently challenges the extrapolation of results outside of individual studies. Second, pain is a complex sensation, and measuring it can be highly challenging. We used standardized mean differences to accommodate differences in tools used to measure pain. However, it is important to understand that the experience of pain can be influenced by a range of factors not necessarily related to the gauge size, which may introduce a challenge in accuracy when measuring pain. One can reasonably speculate that details of different protocols across studies, including the use of topical anesthesia, eye speculum, and povidone-iodine, may influence the results through a confounder effect. Further, aspects not outlined in the studies, such as music in the injection theater, having a neck pillow, patients having their hand held, or experiences in the waiting room/prior to entry to the clinic [31, 43], also have the potential to influence the results, and these aspects were not explored in our review. Controlling for some of these factors in a meaningful manner was not possible in our meta-analysis because of the heterogeneity of the preparation protocols across studies. Third, two studies did not randomize participants, and some concerns were raised across studies because individuals inquiring on the pain experience were aware of the intervention group. Finally, our conclusions are only as good as the studies that provide the data for our analysis. Although nine studies allow for robust analyses, additional studies and more direct comparisons across the groups would provide better certainty around the conclusions.

Conclusions

Our systematic review of the literature and network meta-analysis found a relationship between thinner needle gauge sizes and lower pain experience from intravitreal injections. Our findings are in line with the notions expressed in international guidelines and consensus papers for the administration of intravitreal anti-VEGF injections, and provide clear evidence for decreased pain with the use of 30-gauge or thinner needle sizes. In drug development, we recommend exploring the thinnest feasible gauge size to minimize patient discomfort.