The study of how prescribing philosophies of medical practitioners differ from group to group and from place to place is both interesting and useful. Knowing of these differences will be of value as optometry takes its place in the global medical market.1 More importantly, these differences will be useful in developing standards of practice to improve outcomes. Should there be no emergent consensus of standards, these studies will help to plan future empirical studies to develop evidence-based practices.
In the broader field of medicine, a number of studies exist that study the differences of medical and surgical practice that occur in various geographical areas. The literature on these differences is extensive, both in the different types of medical and surgical procedures considered, the place or country in which these procedures are done, and even in the methods of payment for the services provided. To cite just a few examples, Badeyan et al. catalogued a host of differences in practice that occur in a variety of measures and for a variety of possible reasons between the health care systems of the United States and countries in Europe.2 Using a particular instance in medical practice, Herlitz et al. showed that there are differences in the treatment and outcomes of patients with acute myocardial infarction treated in Minneapolis/St. Paul and Goeteborg, Sweden.3 These differences may be due to historical practice patterns, the structure of the healthcare delivery system, and differences in health care financing. McGlynn et al. compared the appropriateness of a diagnostic and a treatment procedure, coronary angiography and coronary artery bypass graft surgery, between Canada and New York State.4 In this study, the different health care delivery systems did not show significant differences in outcomes.
Spectacle prescribing is both an art and a science, dependent on a host of factors besides just the refractive error. The age of the patient, prior prescription, patient needs, and binocular status are important factors that need to be considered when the doctor goes from gathering data to writing the spectacle prescription. The purpose of this study was to determine how a group of ophthalmologists in a part of Germany would prescribe spectacles for asymptomatic young children of three different ages with different kinds of hyperopic prescriptions, and to compare how they differ from the North American practitioners surveyed by Lyons et al.5
The ophthalmic literature on prescribing for hyperopic children was extensively reviewed in Lyons et al. They found few consistent guidelines for such prescribing in the English literature they reviewed. References in the German ophthalmic literature on prescribing for hyperopic children are equally scarce. Translations of some English language ophthalmic textbooks contain references, and there are a few references by German ophthalmologists.
German ophthalmologists may refer to a number of ophthalmology texts that offer advice on prescribing for hyperopes. Kanski is a widely used ophthalmology text translated from English into German. This text recommends correction of asymptomatic children who have more than 2.5 to 3.0 D of hyperopia, more than 1.5 D of anisometropia, and astigmatism of more than 1.25 D.6 Augustin and Collins is another such translated text. They recommend treatment of over 3 D of hyperopia if there is no strabismus. These authors are, however, much more aggressive if there is amblyopia, recommending full treatment of all ametropia, any astigmatism of more than 1 D and full correction of any anisometropia.7 A German author, Sachsenweger, states that a light to medium grade hyperopia in adolescents does not have to corrected.8 Another German author, Grehn, notes in his text that a young person can easily accommodate through 4 D of hyperopia, but he does so in the context of determining latent hyperopia. He offers no specific recommendations for correcting hyperopia in the age groups under consideration, but he implies that it should be nearly fully corrected.9
Evidential studies have been done that relate to the topics at hand. The pertinent question of efficacy of early intervention with spectacles for hyperopia has been studied by Atkinson10 and by Atkinson et al.11 They found that early spectacle wear in significantly hyperopic children, as measured by photographic and videographic refraction techniques, reduced the incidence of strabismus and amblyopia at 4 years of age. Ingram et al., on the other hand, found the same incidence of squint in those 6-month-old hyperopes who wore glasses as in those who did not.12 Both groups in this study had the same overall incidence of amblyopia. However, of those who were prescribed spectacles, there was a slight improvement in amblyopia if the child consistently wore the spectacles when compared with those who probably or certainly did not wear the spectacles consistently.
A different question is whether or not the early intervention of spectacle prescription and wear in high hyperopes will alter the normal emmetropization process. Atkinson et al. found that such early intervention did not alter the process; the group of 6-month-old hyperopes who wore spectacle decreased their hyperopia at the same rate as those who did not wear spectacles.11 Contrary to these findings were those of Ingram et al., who found that early spectacle prescription did retard, although slightly, the emmetropization process.13
METHODS
In Germany, an Augenoptiker similar to British refracting optician, can prescribe spectacles for patients over 14 years of age. We did not therefore include these practitioners in our study. Neither did we include the very few optometrists practicing in Germany. We surveyed ophthalmologists in an area of Germany to obtain opinions in prescribing for hyperopia, astigmatism, and anisometropia in symptom-free children of ages 6 months, 2 years, and 4 years. These practitioners were asked if and how they would prescribe for a number of hyperopic refractive conditions.
The guidelines of the Department of Clinical Investigation, Walter Reed Army Medical Center, were followed. Requirements for the exemption from the Investigational Review Board were met because no experimental manipulations of human or animal subjects were made in the study. Signature release of patient information was not necessary and there was no identifying information of individual surveys returned by the participating German ophthalmologists.
A sample of names of German ophthalmologists was obtained from a list of ophthalmologists at gelbeseiten.de on the World Wide Web. Some of these medical doctors were in private practice (niedergelassener Augenarzt) and some were affiliated with the teaching hospital at two major medical universities. They all practice ophthalmology in an area known historically as Franconia, now part of the German state of Bavaria. The cities in which they practice were Wuerzburg, Schweinfurt, Nuremburg, and Erlangen. The teaching hospitals in which some of them practice were the Universitaets-Augenklinik, Wuerzburg, and the Augenklinik, Universitaet Erlangen-Nuernberg, two historic and famous medical universities. Permission for use of the surveys in Lyons et al. was obtained from the corresponding author. A native German speaker, fluent in English, translated the English survey of Lyons et al. into German. A different native German speaker, who was also fluent in English, translated them from German back into English. A native English speaker, different from the first, then made only two minor changes and the edited German surveys were sent out. Once we received the returned surveys, the data were entered into a database and analyzed by computer program. The Fisher's exact test was used to test for differences between the data collected from German ophthalmologists and the data collected by Lyons et al.14
Eighteen statistical comparisons using the survey data were computed, all of them using Fisher's exact test. As hypotheses were undirected, two-sided p-values were computed. To control for the effect of multiple comparisons, we applied the Bonferroni correction procedure.15 The individual p-values are compared with a threshold that has been Bonferroni-corrected for the number of comparisons. Thus, the usual threshold of 0.05 becomes a threshold of 0.0028 (5/18 = 0.0028). A p-value that is smaller than 0.0028 is statistically significant.
RESULTS
Of 103 surveys sent out, a total of 45 German ophthalmologists responded (44%). This response rate is about the same as that of the response rate for U.S. ophthalmologists in Lyons et al. The German ophthalmologists who responded did not answer every question, resulting in different totals for the various questions. Only six ophthalmologists of 36 queried at the teaching hospitals responded by returning their surveys. Initially, we intended to compare the responses of ophthalmologists practicing at teaching hospitals affiliated with two famous medical universities with those ophthalmologists in private practice, but because of this low return, we decided to combine the responses of all the German ophthalmologists.
Most German ophthalmologists reported examining 20 or fewer patients younger than 2 years of age per month. The vast majority of U.S. ophthalmologists surveyed in Lyons et al. examined more than 20 patients younger than 2 years of age per month, while most U.S. optometrists in the same study reported examining 10 patients of this age group or less per month. Thus, the German ophthalmologists fall between the U.S. ophthalmologists and the U.S. optometrists in the number of patients <2 years of age they reportedly examine each month (Table 1).
The practitioners were asked how they would prescribe for different levels of bilateral, asymptomatic hyperopia in patients of age 6 months, 2 years, and 4 years (Table 2). Over two-thirds of practitioners of both nationalities queried would prescribe for asymptomatic 6-month olds with more than 5 D of hyperopia. There was no significant difference between the German ophthalmologists and the U.S. optometrists in prescribing for hyperopia in this age group, (Fisher's exact test p = 0.42), but there was considerable difference between the German ophthalmologists and the U.S. ophthalmologists (p < 0.001).
When asked how they would prescribe for asymptomatic 2-year-old hyperopes, the majority of German ophthalmologists would prescribe for 3 D or more of hyperopia. Here again, there was no significant difference between the German ophthalmologists and U.S. optometrists (p = 0.40), but there was significant difference between the German ophthalmologists and U.S. ophthalmologists (p < 0.001).
In the case of 4-year-old patients, a large majority of German ophthalmologists would prescribe for at least 3 D of hyperopia. There was no difference between the German ophthalmologists and the U.S. optometrists (p = 0.05), but there was a significant difference between German ophthalmologists and their U.S. counterparts (p < 0.001). All practitioners would prescribe for lesser amounts of asymptomatic hyperopia as the age of the young patient went up.
The survey queried the practitioners about their philosophy for prescribing the full or less than full (cut) prescriptions for asymptomatic, bilateral 6-month-old hyperopes. Prescribing full or less than full prescriptions of the sphere and/or cylinder portions yielded four possible selections from which the practitioners might choose (Table 3). The majority of German ophthalmologists would prescribe the full cycloplegic amount of hyperopia and less than the full astigmatism portion. In this, they more or less concurred with U.S. ophthalmologists (p = 0.06). However, the German ophthalmologists were significantly different from the U.S. optometrists; the latter would prescribe less than the full amount of both the hyperopic and astigmatic portions of the prescription (p < 0.001).
Of the German ophthalmologists who would prescribe less than the full amount of either or both portions of the prescription for a 6-month-old, bilateral, asymptomatic hyperope, the survey asked about a particular rule of thumb for prescribing. The practitioners were asked if they would prescribe certain fractions of the hyperopic or astigmatic portions of the prescription, or use some other, unspecified, rule of thumb to cut the prescription (Table 4).
The majority of German ophthalmologists would not use the fractional rule of thumb to cut the hyperopic portion of the infant's prescription. The p-value for them and the U.S. ophthalmologists was p = 0.007, and for them and the U.S. optometrists was p = 0.01. There was therefore no statistically significant difference between the German ophthalmologists and the U.S. ophthalmologists or between the German ophthalmologists and the U.S. optometrists in using the fractional rule of thumb for cutting the hyperopic portions of the infant's prescriptions.
Of those few German ophthalmologists who would prescribe a fractional amount of the astigmatic portion of a prescription, the majority would apply the rule of thumb by prescribing from one-half to three-fourths of the astigmatic portion. The p-value when compared with the U.S. ophthalmologists was p = 0.80, and when compared with the U.S. optometrists was p = 0.07. There was no significant statistical difference between these prescribing practices of the German ophthalmologists and those of U.S. ophthalmologists, or between those of the German ophthalmologists and those of the U.S. optometrists.
Finally, the German practitioners were asked how they would prescribe for different amounts of asymptomatic, hyperopic anisometropia for children of different ages (Table 5). For 6-month-old patients, the majority of German ophthalmologists would prescribe for 3 D or more of anisometropia. There was no significant difference between them and both U.S. ophthalmologists (p = 0.50) and U.S. optometrists (p = 0.51). For 2-year olds, the majority of German ophthalmologists would prescribe for 1 D of anisometropia. Their prescribing philosophies were not significantly different from the U.S. ophthalmologists (p = 0.70), or from the U.S. optometrists (p = 0.28). By patient age of 4 years, fully three-fourths of German ophthalmologists would prescribe for 1 D of anisometropia. Here again, they were of one mind with both the U.S. ophthalmologists (p = 0.34) and U.S. optometrists (p = 0.95). The German ophthalmologists generally prescribed for lesser amounts of anisometropia as the age of the patient went up, as did both types of U.S. practitioners.
DISCUSSION
Questions that were conditional on previous responses would necessarily have lower numbers of responses than questions that were not. A few practitioners declined to answer some questions. However, the number of questionnaires returned from German ophthalmologists associated with teaching hospitals at German Universities was very low; only six were returned of 36 sent out for a rate of 19%, compared to a return rate of 67% for the ophthalmologists not associated with teaching hospitals. Perhaps University ophthalmologists examine few children. This was unexpected, because the University teaching hospital clinics were considered referral centers.
Comparing the numbers and percentages of children less than 2 years of age seen by the three different groups of ophthalmic practitioners reveals some remarkable differences, for which there are numerous possible reasons. Structural, educational, and economic differences between the U.S. and German medical systems may be associated with these differences in numbers of patients seen. It is possible that U.S. primary care medical practitioners are more likely to refer to subspecialist ophthalmologists than their German primary care counterparts. Perhaps ancillary medical caregivers in the U.S., such as nurse practitioners, community health nurses, and school nurses are more familiar with pediatric ophthalmologists and therefore refer to them more often than they do to pediatric optometrists. The medical education systems differ for each of the three sets of practitioners and these may play a role in creating these differences. Thus, it may be that the subspecialty education of the U.S. pediatric ophthalmologists trains and prepares them to see very large numbers of patients, while neither the German ophthalmologists nor the U.S. optometrists get such training in high volume practice. It is possible that pediatric ophthalmologists in the U.S. advertise more than pediatric optometrists. The different insurance payment systems of each country may provide for more numerous office visits in the U.S. than in Germany. U.S. optometrists compete with U.S. ophthalmologists to satisfy the demand in this area of the medical market, but in Germany only ophthalmologists fill this niche.
There was little difference between German ophthalmologists and U.S. optometrists in prescribing for asymptomatic bilateral hyperopic patients of all three age groups under study. In some contrast, German ophthalmologists seem to have prescribing philosophies different from the U.S. ophthalmologists in prescribing for bilateral hyperopia at any of the patient ages under consideration. In general, German ophthalmologists consider prescribing for lesser amounts of bilateral hyperopia and do so sooner than the U.S. ophthalmologists.
The previously cited evidential studies address early prescribing for bilateral hyperopia, but unfortunately, they offer little support for either philosophy. These studies are few in number and do not offer a consistent indication for the efficacy of early prescription for very young, bilateral, asymptomatic hyperopes. Similarly, the two studies on the effects on emmetropization of early prescription are not in agreement.
If the evidential studies provide no support for one side or the other in this controversy, what might account for the differences for these different prescribing philosophies? The possible causes offered are highly speculative. The U.S. optometrists and German ophthalmologists saw fewer patients per month than do the U.S. ophthalmologists, so the two former groups may have more time per patient than the U.S. ophthalmologists. Perhaps the extra time these two groups have allows them to do more clinical tests that convince them to prescribe earlier than the U.S. ophthalmologists, although we have no data about the number or types of tests used by each set of practitioners. Perhaps the ready availability of eyewear coverage for medical reasons by German health insurance system, the Krankenkasse, may make those practitioners more likely to prescribe spectacles.
Another likely reason for these differences in prescribing philosophy, however, concerns the niche these groups occupy in their medical care systems. The German ophthalmologists are certainly the only primary care practitioners for all three age groups under study. The U.S. optometrists in Lyons et al., despite their distinction as pediatric optometrists, probably functioned as primary care givers, while the U.S. ophthalmologists in that study were subspecialist secondary and tertiary care practitioners. Thus, the early prescribers were primary care givers, while the late prescribers were not. These structural differences in the ophthalmic health care systems could be associated with the differences in prescribing philosophies.
Cutting the prescription in hyperopia or astigmatism corrections is a rule of thumb used by some practitioners to increase the acceptance of the prescription. It is sometimes used in older patients who are able to give more feedback about the prescription, but its use for children as young as those under consideration may not be valid. Fractionally cutting the prescription in an objective examination, which is certainly what such an examination of a 6-month-old patient is, seems to be a rule of thumb favored by a majority of U.S. optometrists for both the hyperopic and astigmatic portions of the prescription according to Lyons et al.5 This particular rule of thumb was not a favorite of the German ophthalmologists for either portion of the prescription.
The Other category was selected by 71% of German practitioners responding to the question about fractional prescriptions for hyperopia. This may indicate that this particular rule of thumb of fractionally cutting a prescription is not widely used among the German ophthalmologists. Other rules of thumb, perhaps involving linear decreases instead of fractional ones may be preferred. For instance, there is a rule of thumb specifically mentioned in Grehn, recommending that a prescriber cut the cycloplegic refraction by 1/2 D for strabismic patients.9 Although unrelated to the conditions present in the survey questions we asked, it may be that German prescribers default to this or some other rule of thumb in the absence of any satisfying the specific conditions asked about in the surveys.
The issue of rule of thumb deserves some discussion. Although they have not been widely studied, some attention was given them in the context of artificial intelligence.16,17 In the context of the current study, only a few of the prescribers liked the two rules of thumb presented to them in the questionnaire. Before experts can decide on which rules of thumb to use, they should have access to a large set of such prescribing rule from which to choose. For this, a web-based clearinghouse of prescribing rules could be established. A large collection of prescribing rules could be assembled from expert and experienced submitters to the site. From among this large set of rules, various selection mechanisms, such as voting, expert system analysis, etc., could be used to determine the most popular set of rules of thumb. These could then be empirically tested using standard techniques.
The treatment of anisometropia is an area where there appears to be considerable agreement between the German ophthalmologists and the U.S. practitioners, for all age groups of patients. This similarity was noted by Lyons et al. for the two different types of U.S. practitioners. The majority of all three groups of prescribers believe in correcting anisometropia early and at low levels, certainly at 3 D of anisometropia for the 6-month-old patients and at 1 D for the older patients. This remarkable unanimity is consistent with studies relating to anisometropia and the genesis of amblyopia, such as that of Weakly.18
The current survey samples, as did that of Lyons et al., attitudes and not behaviors. Clearly, more evidence-based studies need to be done on the efficacy and safety of early prescribing for a young asymptomatic bilateral hyperope, based on clinical outcomes. These would determine the range of ages at which hyperopic prescriptions should be started and for what amounts of hyperopia these prescriptions need to be given. Putting the opinions of the three groups of eye care practitioners to the empirical test, we might arrive at rules of thumb based upon specific age groups for hyperopic prescriptions. Such evidence-based studies in the age groups under consideration may require considerable time, thus increasing their cost and difficulty. In the absence of these evidence-based studies, we are left with measuring attitudes, intentions and prescribing philosophies of experts to standardize the medical practices.
ACKNOWLEDGEMENTS
The views expressed in this article are those of the authors and do not necessarily reflect the positions of the Julius-Maximilians Universitaet, The Interdisziplinaeres Zentrum fuer klinische Forschung, the Zentrale fuer klinische Studien, nor the Department of the U.S. Army.
Ellis Madsen
951 Sierra Madre Drive
Salinas, California 93901
e-mail: ellis.m.madsen@us.army.mil
REFERENCES
1. Di Stefano AF, Chappell R, Smith D, Wallis N, Lawless T, Penisten D, Norwood B, Naidoo K. The globalization of optometry: challenges and opportunities in the new millennium-part I. Optometry 2004;75:341-6.
2. Badeyan G, Foulon D, Gottely J, Gottely P, Pauriche P. A comparison of health in Europe and America. Institut National de la Statistique et des Études Économiques Donnees Socials (INSEE), 1999 ed. Available at:
http://www.insee.fr/en/ffc/docs_ffc/ds9936.html. Accessed November 27, 2006.
3. Herlitz J, McGovern P, Dellborg M, Karlsson T, Duval S, Karlson BW, Lee S, Luepker RV. Comparison of treatment and outcomes for patients with acute myocardial infarction in Minneapolis/St. Paul, Minnesota, and Goteborg, Sweden. Am Heart J 2003;146:1023-9.
4. McGlynn EA, Naylor CD, Anderson GM, Leape LL, Park RE, Hilborne LH, Bernstein SJ, Goldman BS, Armstrong PW, Keesey JW, et al. Comparison of the appropriateness of coronary angiography and coronary artery bypass graft surgery between Canada and New York State. JAMA 1994;272:934-40.
5. Lyons SA, Jones LA, Walline JJ, Bartolone AG, Carlson NB, Kattouf V, Harris M, Moore B, Mutti DO, Twelker JD. A survey of clinical prescribing philosophies for hyperopia. Optom Vis Sci 2004;81:233-7.
6. Kanski JJ. Lehrbuch der klinischen Ophthalmologie, vol 2: Auflage. Stuttgart: Georg Thieme; 1996.
7. Augustin AJ, Collins JF. Augenheilkunde, vol 2: Ausgabe. Berlin: Springer Verlag; 2001.
8. Sachsenweger M, Klauss V, Hasemann J, Ugi I. Augenheilkunde, vol 2: Ausgabe. Stuttgart: Georg Thieme; 2003.
9. Grehn F. Augenheilkund. Berlin: Springer Verlag; 2003.
10. Atkinson J. Infant vision screening: prediction and prevention of strabismus and amblyopia from refractive screening in the Cambridge Photorefraction Program. In: Simons K, ed. Early Visual Development: Normal and Abnormal. New York: Oxford University Press; 1993:335-48.
11. Atkinson J, Braddick O, Robier B, Anker S, Ehrlich D, King J, Watson P, Moore A. Two infant vision screening programmes: prediction and prevention of strabismus and amblyopia from photo- and videorefractive screening. Eye 1996;10(Pt 2):189-98.
12. Ingram RM, Arnold PE, Dally S, Lucas J. Results of a randomised trial of treating abnormal hypermetropia from the age of 6 months. Br J Ophthalmol 1990;74:158-9.
13. Ingram RM, Gill LE, Lambert TW. Effect of spectacles on changes of spherical hypermetropia in infants who did, and did not, have strabismus. Br J Ophthalmol 2000;84:324-6.
16. Madsen EM, Reinke AR, Rehrs MH, Yolton RL. Exploring the optometric application of expert computer systems: refractive error correction. J Am Optom Assoc 1991;62:621-9.
17. Madsen EM, Reinke AR, Fehrs MH, Yolton RL. Applications of expert computer systems. J Am Optom Assoc 1991;62:116-22.
18. Weakley DR. The association between anisometropia, amblyopia, and binocularity in the absence of strabismus. Trans Am Ophthalmol Soc 1999;97:987-1021.