We thank Professors Kommerell and Bach for their interest in our article.1 Mallett advocated the use of a trial frame with his test as it allows a normal head posture and visual field.2 Mallett advocated using small step sizes (1D horizontally);2 “gradually increasing the strength until the slip disappears – never the reverse procedure”;3 and “between changes of prisms or spheres the patient should read two or three lines of print surrounding the target”.3 These instructions depart markedly from the procedure adopted by Kommerell and Bach,4 where the participant continuously adjusts a Risley prism, on several occasions starting with 10Δ, and is asked to “play a little”(with the prism power). We agree with Kommerell and Bach, their method did not aim at the smallest power of the prisms and may not be appropriate for prescribing.

We understand how a consideration of natural viewing led Kommerell and Bach to the interesting approach of self-selected prism.4 The assumption behind this test seems to be that following a period of self-adjustment of a Risley prism, a subject's selection of the prism power they find most relaxing may be therapeutically helpful. We question this hypothesis for several reasons, most notably that the self-selected prism fluctuates considerably from one day to another4 and subjects may select the strongest prism they can tolerate, not the weakest.

We accept that Kommerell and Bach's avoidance of Nonius markers is a step towards normal viewing conditions, but we suggest that various aspects of their experimental design, including the participant adjustment of Risley prisms, takes this approach several steps further away from normal viewing conditions. In support of this we note, the prism powers found by Kommerell and Bach4 are generally more than double those typically obtained with the Mallett unit, when used as recommended.5 We made the comment in our manuscript advising against using self-adjusted Risley prisms because of concerns that if clinicians use the Mallett unit in this way they could inappropriately over-prescribe prisms, both in the proportion of patients to whom prisms are prescribed and in the magnitude of prism.

Practitioners who use the Mallett unit tend to only prescribe prisms of low power to a small minority of patients with significant symptoms associated with visual tasks when other treatment approaches are unsuitable.6 For example, NHS statistics for Scotland indicate ∼1% of NHS funded lenses supplied by community optometrists include a prism.7 We are not surprised8 that these data on actual practice differ markedly from surveys of practitioners’ choices given hypothetical prescribing scenarios.9

There is experimental evidence supporting the use of the Mallett unit for detecting symptomatic heterophoria10 and for prescribing.1112 As Kommerell and Bach note, there is a fairly large body of research that has used the Mallett unit following the designer's instructions.6 Our concern is that, since the test results are sensitive to differences in instructions13 and test design,1 if clinicians use the test in an unintended way this could lead to unintended consequences, including the over-prescribing of prisms.