The indications for screening are generally accepted [15].

However, a controversy arises over the interpretation of these guidelines. Part of the controversy has to do with what is the intent of screening. If the primary intent is to detect toxic dosing, as has been a theme of this book, then the first criterion is met, because at least 12.8 % of patients taking 4AQs are overdosed (see Chap. 7) [12, 16]. But if the primary intent is to detect retinopathy at an early stage, then it may not be satisfied, because the prevalence of 4AQR has not been established and some estimates of its prevalence are far less than 1 % (see Chap. 5). The best estimate for prevalence of 4AQR probably lies in the range from 0.1 to 5 % depending on drug (higher for chloroquine) and context. If it is close to 0.1 %, then the first criterion has probably not been satisfied. If it lies closer to 5 %, then it probably has been satisfied. Where the decision swings from probably not justified to probably justified is a matter of opinion and resources.

The second criterion is noncontroversial—all agree that advanced 4AQR is serious. The third is controversial, and depends again on the intent of screening. If it is to detect toxic dosing, then there is effective treatment, namely to correct dosing based on ideal body weight (IBW). If it is to detect early retinopathy, then the matter is unclear, because there is inconsistent evidence that retinopathy is reversible (see Chap. 6) [17].

In the debate on the intent of screening, one side holds that the main purpose is to detect retinopathy at an early stage and prevent its progression to a more advanced stage [18–21]. This point of view emphasizes that preventing retinopathy may not be possible [6, 20, 22]. The other view contends that prevention of retinopathy is the goal, which requires a shift in emphasis to the detection of toxic dosing [10, 12, 23–26]. The author’s view is that not enough attention has been given to daily dosing. For example, in 2013 Schwartz and Mieler wrote that “Earlier recommendations emphasized dosing by weight,” as though this point of view is now antiquated. However, such recommendations are forward-looking, because overdosing is common [12, 27, 28]. By reframing the purpose of screening as primarily detection of toxic dosing, with a subsidiary goal of detecting subclinical retinopathy, the tenor of the discussion about screening changes [4, 10, 29, 30].

Although detection of toxic dosing is the most important goal of screening, the lesser goal of detecting retinopathy early does deserve attention. Because 4AQR routinely develops before symptoms are noticed, it follows that screening asymptomatic subjects taking 4AQs will be necessary to detect most cases of retinopathy [31]. Because there is no evidence that retinopathy is reversible to a clinically important extent once it has advanced to the stage of funduscopic changes, a pertinent question is whether detection of retinopathy at the stage of premaculopathy is effective at preventing retinopathy. In Bernstein review of 1992, there were no cases in which premaculopathy had been followed and evolved into more advanced retinopathy [32]. However, in Chap. 6 a well-documented case was presented (Fig. 6.​5), showing that such progression can occur. The evidence is also strong that progression to advanced retinopathy occurs less frequently if retinopathy is detected early and the drug is stopped (see Chap. 6). Cessation of the drug at the premaculopathy stage is associated with a good prognosis [33].

There are other, less important, reasons to screen and reasons that offer incentives to screening. These include a desire to limit medicolegal liability, to respond to patient insistence, to remind patients of the possibility of toxicity so that the drug may be considered for discontinuation if not needed any longer, and to garner remuneration associated with screening [7, 12, 34–36]. These purposes deserve notice, because even if an analysis of the evidence leads to a conclusion at odds with screening, there is a probability that a culture of screening will continue [7].

The ophthalmic literature is inconsistent on the value of baseline screening [4, 13, 34, 37, 38]. For example, Marmor argues for uniformly obtaining baseline visual fields in one place and in another states, “I often omit baseline visual fields” [6, 22]. The rationale for baseline screening is that preexisting maculopathy is a risk factor for 4AQR and therefore needs to be established. Moreover, establishing the presence of a normal baseline makes subsequent determination of maculopathy more significant [39, 40].

When to perform baseline screening is an issue. Bernstein has recommended that it be performed within 4–8 weeks of initiating therapy, but there is no data to suggest that another window might not be acceptable. Others have recommended that the baseline examination be done after 6–12 months of treatment [14, 32]. Although one might argue for a baseline examination before beginning therapy, a practice of doing it within 4–8 weeks has the advantage of not screening those patients who will be intolerant of 4AQs due to gastrointestinal upset or another problem (see Chap. 3). Table 9.1captures the spectrum of published recommendations on baseline screening for 4AQR.

As with recommendations regarding the need for a baseline screening visit, those for follow-up screenings are inconsistent (Table 9.1). The American Academy of Ophthalmology guidelines of 2011 and others suggest a second screening visit after 4–6 years, assuming that no risk factors for retinopathy are present [6, 14, 62, 67]. Despite these recommendations, in practice few clinicians adhere to the advice, because of fears that patients will be lost to follow-up during the prolonged gap [12, 34]. Instead, the prevailing practice is to screen at least yearly [12, 34]. Despite the lack of evidence supporting it, screening every 6 months is also widespread [12, 68].

A rational approach to the frequency of screening should include addressing the question of how fast a patient can go from no damage to visually disabling damage. The answer is not known and perhaps is unknowable since it will depend on the perspicacity of the screener. For example, the patient shown in Fig. 8.​13 received recommended annual 10-2 VFs, but developed 4AQR because the ophthalmologist did not recognize the characteristic paracentral scotoma of the condition.

Many authors place a screening responsibility on the treating physician [4, 32, 39]. There is no objective way to measure the effectiveness of these admonitions, but the continuing occurrence of retinopathy and overdosed patients suggests that previous efforts have not been effective. Some treating physicians are adamant that ophthalmologic screening is needed rather than screening by the treating physician [25]. This appeals to common sense. Imagining the shoe on the other foot, the author shudders to think of ophthalmologists screening for heart murmurs. On the other hand, some rheumatologists view a requirement to involve ophthalmologists in screening as a burden. If this need removed, that use of 4AQs might increase [39].

The screening methods recommended by authors have changed over the years, generally in the direction of more recommended tests in addition to clinical examination (Table 9.1). For example, in 1978 Dubois stated that screening should include clinical examination only and that routine use of visual fields, color vision testing, or any other ancillary modality was unnecessary. In 1992, visual field testing was still considered unnecessary for routine screening [69]. By 2002 standard automated perimetry (SAP) became a universally recommended screening modality [18]. In 2011 at least one of mfERG, SD-OCT, and fundus autofluorescence (FAF) was added to the list of routinely recommended ancillary tests, when these modalities were available [6].

Although governing bodies have issued guidelines [6], there remains no internationally accepted method of screening, any consensus as to which test is most useful, or in what order tests should be obtained [70, 71]. Experts have changed their opinions over time. For example, Easterbrook opined in 1993 that 10-2 VF testing was not indicated on a routine basis in patients taking hydroxychloroquine, but only to confirm a positive Amsler grid test [49, 50]. However, in 2002 he wrote that 10-2 VF testing should be routinely obtained in such patients [18]. In 2002, Marmor wrote that mf ERG should be optional, but by 2011 he changed his mind and wrote that it (or SD-OCT or FAF) should be obtained routinely if available [6]. He changed his mind again in 2013, stating “I often omit baseline fields and almost never order baseline mf ERG testing…I add mf ERG at some point between 5 and 10 years and usually order it only every few years or when something is suspicious on other tests” [22]. No level 1 evidence exists to address issues regarding 4AQR screening.

The single most important component of the clinical examination is the determination of the patient’s height and IBW (see Chap. 7) [10]. This allows the clinician to detect the presence of toxic dosing and to correct it. The remainder of the clinical examination is not sufficiently sensitive to be of great value. It is universally agreed that fundus changes of 4AQR are late (see Chap. 6), and therefore unhelpful for detecting retinopathy at a time when intervention can prevent damage. For this reason, ancillary testing has always been employed (see Chap. 8). The characteristics of the ancillary tests become important in choosing which tests to use. The ideal test should be both sensitive and specific. Unfortunately, in clinical practice, the more sensitive a test is, the less specific it is, and vice versa.