The influence of altitude on the differences between Goldmann tonometry and Pascal dynamic contour tonometry: An ecological meta-analysis

Abstract

Purpose:

A meta-analysis found that including atmospheric pressure as altitude in generalized linear models reveals higher differences between Goldmann tonometry and Pascal dynamic contour tonometry at higher altitudes, with the difference increasing in thinner corneas. To examine the difference in intraocular pressure (IOP) measurements by using Goldman applanation tonometry (GAT) and dynamic contour tonometer (DCT) tonometry in published literature and determine the influence of central corneal thickness (CCT), age, and altitude on that difference.

Methods:

Articles that compare GAT and DCT were selected for an extensive literature review, and the location and altitude of the research centers were found online. CCT and age were analyzed as covariates, when available.

Results:

A total of 157 studies including 24,211 eyes of 20,214 patients were included in the study. The results showed that the difference between DCT and GAT was higher at higher altitudes above sea level and increased with thinner corneas. However, the results were different in eyes with corneal transplants, where altitude and CCT had less influence, and in those post-refractive surgery where age was found to influence the difference. Theoretical correction formulas using altitude, CCT, and age were derived from this meta-analysis, but their accuracy and usefulness in clinical practice need validation.

Conclusion:

The findings suggest that there is a higher risk of underestimating IOP when the Goldmann tonometer is used at a higher altitude, particularly in eyes with glaucoma, thinner corneas, or corneal refractive surgery. Further research is needed to validate the accuracy of the correction formulas derived from this meta-analysis in clinical practice.

Many studies have shown an important correlation of thinner central corneal thickness (CCT) and lower intraocular pressure (IOP) measurements with Goldmann applanation tonometry (GAT). However, despite the recognition of the correlation between corneal thickness and IOP, most correction formulas used to correct these inaccuracies have proven to be unreliable for clinical practice.[1,2,3,4,5,6] In the Ocular Hypertension Study (OHTS), thinner corneas were found to be an important risk factor for patients with ocular hypertension (OHT) developing glaucoma; however, the use of five different correction formulas for GAT using CCT failed to improve the prediction models.[7] The authors suggest that the influence of CCT on glaucoma development is not solely due to IOP measurement error but is also a reflection of other factors that play a role in glaucoma development. The quest for measuring “real” IOP without a direct intraocular measurement via manometry has led to the development of several tonometers, with the dynamic contour tonometer (DCT), also known as Pascal® (in honor of Blaise Pascal), being one of the most promising in the attempt to overcome the inherent drawbacks of GAT.

DCT compensates for the variability of intraocular measurements caused by CCT, corneal hysteresis, and other biomechanical properties.[8] In several studies, DCT has been shown to perform better than GAT, with fewer inaccuracies related to corneal characteristics.[9,10,11,12,13,14,15] Furthermore, the Pascal tonometer is a close equivalent to manometric measurements, making it a potential candidate to replace GAT as the new gold standard in IOP measurement.[16,17,18] Most studies report that IOP readings obtained by DCT are significantly higher than those obtained by GAT, with most authors not agreeing on the magnitude of this difference nor its association with the biomechanical properties of the cornea.[9,10,11,12,13,14,15,16,17,18,19]

In previous studies, we have reported that the difference in IOP measurements between sea level and 2234 m above sea level in the same group of subjects was found to be significantly greater using GAT than using DCT, and this difference was found to be replicable in a hyperbaric chamber.[20,21]

This study aims to evaluate the difference in IOP measurements between DCT and GAT and estimate the effect of CCT, age, and altitude on that difference.

Methods

Literature search strategy

A comprehensive literature review was conducted using several databases, including PubMed, Scielo, AAO.org, LILACS, Google Scholar, and LinkedIn. Articles, theses, and reports regarding studies on comparison between Goldmann and Pascal tonometers were selected when paired data were available. Languages included in the search were English, Spanish, Portuguese, French, and German. The authors attempted to obtain missing data by contacting the corresponding authors of the selected studies. The keywords were (DCT OR dynamic contour tonometry OR Pascal tonometry) AND (Goldmann tonometry OR GAT OR Applanation tonometry). We included papers that reported mean GAT and DCT and excluded those with incomplete data, those that used the same patients for different publications, cadaver eyes, those that did not compare tonometers in the same patients, other languages, review papers, and case reports.

Variables obtained from the resulting studies were mean GAT and DCT IOP, mean age, number of eyes, number of individuals, diagnosis, CCT (regardless of the technique), and altitude (in m). When data from more than one observer was reported, we used the data from the observer with the lowest standard deviation (SD) with DCT first and then GAT or the mean measurement of all observers, if reported.

The location of the research centers participating in the studies was determined using Google Maps, and the altitude was calculated using the Daftlogic app (https://www.daftlogic.com/sandbox-google-maps-find-altitude.htm). In cases where a study was multicentric and had different altitudes at each location, a weighted mean altitude was calculated when the authors provided the proportion of patients at each location; otherwise, a simple mean altitude was used if such information was not available.

The eyes in the selected studies were divided into five diagnostic groups: healthy eyes, glaucoma eyes, keratoconus, corneal transplant eyes, and refractive surgery eyes. Each group was analyzed separately to ensure that the results were specific to each diagnosis.

The healthy eyes group included eyes labeled as “normal,” “healthy,” and eyes before refractive surgery. The glaucoma eyes group included eyes diagnosed with primary open-angle glaucoma, normal-tension glaucoma, ocular hypertension, glaucoma suspects, congenital glaucoma, or post-glaucoma surgery. The keratoconus eyes group included all degrees of the condition and were treated as separate groups when reported as such. The corneal transplant eyes group included eyes that underwent Descemet stripping endothelial keratoplasty or penetrating keratoplasty. The refractive surgery eyes group included eyes that underwent LASIK, LASEK, PRK, or intra-LASIK.

The mean values included in each diagnostic group were weighted by the number of eyes in the original subset from each paper to ensure that the results were representative of the entire population. A multivariable generalized linear model was constructed using DCT-GAT as the dependent variable, altitude (as a surrogate of atmospheric pressure for ease of use) as the independent variable, and CCT and age as covariates. A linear regression analysis was performed to investigate the relationship between age and CCT When a significant correlation was found (CCT is usually reported thinner in older Glaucoma patients, whereas older Normal patients usually have thicker corneas), only CCT was used as the covariate in the final model as it has been extensively studied and its impact on GAT is well established.

Results

The full text of 270 studies was indexed or available online and published between January 2003 and March 17, 2016. A total of 157 papers and theses comparing Goldmann tonometry and Pascal tonometry on different populations were reviewed and included in the study. A total of 24,211 eyes from 20,214 subjects were assessed and included in 276 subject groups [Table 1].

Table 1.

Distribution of study groups, eyes, and patients according to diagnostic group

Type of eyes		# of groups		# of eyes		# patients
Healthy		105		10811		9252
Glaucoma		107		10036		8283
Keratoconus		14		775		635
Transplants		14		426		416
Refractive		36		2163		1628
Total		276		24211		20214

The altitude of the locations within the studies ranged from 3 m to 2252 m above sea level. The linear regression of CCT and age was performed, and due to a significant correlation, only CCT was used in the predictive models for healthy, glaucoma, keratoconus, and corneal transplant eyes. Age was found to be correlated with thinner corneas in glaucoma patients and with thicker corneas in the other three groups. For the refractive surgery group, both CCT and age were used for analysis as no significant correlation was found between the two variables [Table 2].

Table 2.

Linear regression of CCT and age by diagnostic group. CCT=Central corneal thickness

Type of Eyes		R		Constant (microns of CCT)		B coefficient (change in microns per year of age)		P
Healthy		0.166		539		0.152		<0.001
Glaucoma		0.138		561		−0.294		<0.001
Keratoconus		0.450		424		1.277		<0.001
Transplants		0.625		456		2.662		<0.001
Refractive		0.028		517		−0.215		0.236

Multivariable generalized linear models were constructed and showed a statistically significant increase in the difference between DCT and GAT with higher altitude in healthy, glaucoma, and keratoconus eyes. A formula was developed for each study group based on this data and is presented in Table 3 and Fig. 1a-c.

Table 3.

Linear regression formulas from multivariable generalized linear analysis for estimating DCT-GAT (delta) by diagnostic group

Type of Eyes		Formula from linear model		P		Plain language interpretation
Healthy [Fig. 1a]		Delta=13.56+0.71*Altitude -0.022*CCT		<0.001		increase of 1 mmHg for every 710 m of altitude. decrease of 1 mmHg for every additional 45 μm of CCT.
Glaucoma [Fig. 1b]		Delta=10.44+0.536*Altitude -0.015*CCT		<0.001		increase of 1 mmHg for every 536 m of altitude. decrease of 1 mmHg for every 66 μm of CCT.
Keratoconus [Fig. 1c]		Delta=18.54+1.197*Altitude -0.034*CCT		<0.001		increase of 1 mmHg for every 1197 m. decrease of 1 mmHg for every 29 μm.
Transplants [Fig. 1d]		Delta = -2.483+2.483*Altitude +0.007*CCT		<0.001		increase of 1 mmHg for every 2480 m. increase of 1 mmHg for every 143 μm.
Refractive [Fig. 2a and b]		Delta=13.47-0.373*Altitude -0.028*CCT +0.097*age		<0.001		decrease of 1 mmHg for every 373 m of altitude. decrease of 1 mmHg per 36 μm of CCT. an increase of 1 mmHg for every decade of age.

The result for delta should be added to GAT to estimate DCT. Altitude should be in km, CCT in μm, and age in years. DCT=Dynamic contour tonometry, GAT=Goldmann applanation tonometry, CCT=Central corneal thickness

Figure 1.

Comparison of the delta between DCT and GAT with CCT and altitude across different eye conditions. (a) Healthy eyes: This graph illustrates the relationship between the difference between DCT and GAT (delta) with CCT and altitude in normal eyes. It shows that as the CCT decreases and altitude increases, the delta between DCT and GAT also increases. (b) Glaucoma eyes: Similar to normal eyes, the delta increases as the CCT decreases and altitude increases. However, the slope of the effect of CCT and altitude is not as steep as in normal eyes. (c) Keratoconus eyes: The delta increases as the CCT decreases and altitude increases; however, the slope of the effect of CCT and altitude is steeper than in normal and glaucoma eyes. (d) Corneal transplants: This graph illustrates a unique relationship between the delta with CCT and altitude in corneal transplant eyes. The delta between DCT and GAT is higher at sea level compared to higher altitudes, which is the inverse of what is seen in normal, glaucoma, and keratoconus eyes

Corneal transplant eyes showed different results, being less affected as an increase in the difference between DCT and GAT of 1 mmHg was found every 2480 m, coupled with an increase in the difference per every 143-µm increase in CCT (P < 0.001, Table 3, Fig. 1d).

For eyes with corneal refractive procedures, the model included age and showed a significant decrease in the difference between DCT and GAT of 1 mmHg per every 373 m of altitude, a decrease of 1 mmHg per every 36-µm increase of CCT, and an increase of 1 mmHg per decade of age (P < 0.001, Table 3, Fig. 2a and b).

Figure 2.

Comparison of the delta between DCT and GAT with CCT and age in refractive surgery eyes at different altitudes. (a) 0-km altitude: The delta increases as the CCT decreases and age increases. (b) 3-km altitude: The delta between DCT and GAT increases as the CCT decreases and age increases; however, the increase in altitude reduces the absolute difference between the two measurements

A posthoc analysis was performed to calculate the variation of CCT and IOP per kilometer of altitude. Keratoconus eyes encountered the highest variation in CCT (−14 µm), while Glaucoma eyes showed the least variation (+1.5 µm) [Table 4].

Table 4.

Changes in CCT and IOP per km of altitude among the five groups of eyes. The changes in CCT were derived from linear regression, and those in IOP are the reciprocal of the values in Table 3

Type of Eyes		Change in CCT per km of altitude (in μm)		P		Change in IOP difference per km of altitude (in mmHg)		P
Healthy		+2		<0.001		1.4		<0.001
Glaucoma		+1.5		<0.001		1.86		<0.001
Keratoconus		−14		<0.001		0.83		<0.001
Transplants		−16		0.211		0.40		<0.001
Refractive		2.9		0.116		2.68		<0.001

Discussion

The results of our study highlight the importance of considering altitude when interpreting tonometry measurements, particularly when comparing GAT and DCT, confirming previous experimental data on normal eyes.[20,21] It is well established that CCT is a key factor in the accuracy of IOP measurements and that it can vary with age and other factors. Our study supports these findings and adds altitude to the list of important variables that can affect IOP measurements. By incorporating all three variables into our models, we developed a more comprehensive understanding of how altitude can influence IOP measurements.

One interesting aspect of our study is the finding that corneal transplant eyes were less affected by altitude than other groups. This is in contrast to the results for healthy eyes, glaucoma patients, and those with keratoconus, who showed a statistically significant increase in the difference between GAT and DCT with higher altitudes. The reasons for this difference are not entirely clear; however, it may reflect differences in corneal biomechanics between transplant eyes and other groups.

Age has been reported to act as a surrogate for corneal biomechanics and is incorporated as such in the development of the Corvis ST (Oculus ®, Wetzlar, Germany) biomechanically corrected IOP (bIOP), jointly with CCT.[22,23] Interestingly, bIOP seems to measure IOP closer to real IOP in Joda et al.’s finite element model eye and behaves accordingly in real eyes.[22] Our study agrees with both significant variables and adds altitude as a third significant variable. As such, altitude might also influence the behavior of air-puff tonometers as they also use corneal applanation to estimate IOP.

Changes in CCT with altitude have been found in different studies and can account for the discrepancy in IOP measures, both in chronic and acute circumstances.[24,25,26,27] However, in our study, the amount of cornea thickening was minimal, a mere micron per kilometer in altitude, and it did not correlate with the magnitude of the difference in tonometry values. Of note, we do not discard that other biomechanical properties of the cornea, such as corneal hysteresis, might influence this difference; further research using the appropriate equipment is therefore needed.

Our study also revealed that the keratoconus group had a significant change in CCT with altitude, with a decrease of −14 µm/km. This is a novel finding that has not been reported in previous studies, and further research is needed to explore the implications of this relationship regarding a possible role in the pathogenesis, increased prevalence, disease degree, or need for corneal transplant of keratoconus patients at higher altitudes.

An intriguing finding of this study is the differing linear correlations between CCT and age in normal and glaucoma eyes [Table 2]. This has not been previously described and creates a new source of bias. In normal eyes, as CCT increases with age, this increase is interpreted as an overestimation by GAT of the true IOPs by using traditional correction formulas. However, in glaucoma eyes, older age is linked to thinner corneas, which is interpreted as an underestimation of the actual IOPs. This disparity can be attributed to the corneal effects of topical hypotensive medications, such as prostaglandins, which were not accounted for in this study.[28]

The results of the formulas developed in this study highlight the significance of CCT in GAT measurements. When altitude is close to sea level, the difference between GAT and DCT can largely be attributed to CCT in most eyes, and age in eyes post refractive surgery.

The relationship between the increasing DCT-GAT difference and CCT in corneal transplant eyes can be due to subclinical corneal edema rather than a naturally thick cornea. This may affect the biomechanical behavior of this group of subjects.

Refractive eyes require special consideration. Although the behavior of altitude and CCT may seem counterintuitive, the amount of ablation in these eyes is proportional to the presurgical refraction, making the final thickness independent of age. This independence is why CCT and age are not related in this group of eyes. Because CCT and corneal curvature are both modified by refractive procedures, we hypothesize that lower atmospheric pressures might make these corneas a bit steeper, needing a bit more force to applanate, coupled with less stiffness from the induced thinning in younger subjects to more stiffness despite induced thinning in older individuals. Further investigation using both clinical and finite model data is needed to fully understand this phenomenon.

An advantage of our approach, using DCT as a substitute for manometry, is that all Pascal tonometers are manufactured by the same provider worldwide. However, a limitation is that various GAT tonometers are available, which differ in materials, tips, calibration systems, etc., and that GAT can be performed with or without fluorescein. Reported values have shown significant differences between manufacturers despite proper baseline calibration.[29,30,31] Information regarding GAT manufacturers was not retrievable from individual papers for our analysis, and this may have introduced additional bias into the results.

The possible explanation for the effects of atmospheric pressure on GAT is mainly due to physics. GAT and other tonometers that induce a change in the shape of the cornea induce more corneal deformation when the IOP is lower and less deformation when the IOP is higher. Dawson et al.,[32] in “Adler’s Physiology of the Eye,” mention three outward stressors on the cornea: IOP, atmospheric pressure, and the eyelids’ resting pressure. These outward stressors increase resistance to corneal deformation; thus, GAT reflects a higher measurement if any of the three variables is increased. Atmospheric pressure is 760 mmHg at sea level and gradually decreases at higher altitudes. This means that lower atmospheric pressures permit GAT and other force-based tonometers to cause greater corneal deformation and, therefore, show a lower measurement even if real or manometric IOP is the same in the same individual.

There are many clinical implications of this physical phenomenon. As an example, glaucoma patients travel to different cities to obtain second opinions. Countries such as Colombia and Mexico have their capital cities, Bogota and Mexico City, at quite high altitudes. These cities tend to have a higher population, thus requiring a higher number of ophthalmologists to attend. Because they are also the political centers, they tend to attract the most renowned teaching hospitals and glaucoma specialists in each country. Glaucoma patients who might have had elevated IOP at a sea-level city might be erroneously classified as normal-tension glaucoma or even as “normal” patients based on IOP alone. The opposite might occur in countries such as India and Peru, with capital cities closer to sea level, or in a decentralized country such as the United States, where many large teaching hospitals are at sea level in cities such as Miami, Boston, San Francisco, and Los Angeles.

Another important clinical implication is regarding multicenter studies. Their design implies several advantages, such as greater and faster recruitment of patients, diverse population coverage, and increased generalizability. However, it also implies that site selection should be systematic so that intersite variability is minimized.[33] This also means that all sites should follow the same protocol for recruitment, measurements, and diagnostic machines such as OCT, etc., In glaucoma protocols, this means using the same type of tonometry, and it also means using GAT as the time-proven, clinically accepted standard. If the study design only includes sites near sea-level cities, GAT and CCT will be sufficient and the results will be homogeneous. However, if some patients are recruited in higher cities, such as Denver (1608 m), Mexico City (2240 m), or some cities in India above 3000 m (e.g. Kashmir), GAT measurements will have significant intersite variability. It is even possible that reports of surgical procedures at higher altitudes might have lower mean IOP as compared to previous literature not because there are better surgeons at higher altitudes or because their technical variation works better but simply because GAT will yield a systematically lower reading in all patients. When using GAT in a surgical report, “Success” is usually defined as the final IOP of 10–21 mmHg using GAT, making the use of absolute IOP levels introduce additional bias into the analysis. When reports add the use of percent IOP reduction, the risk of bias is reduced.[34]

In addition to atmospheric pressure, we have to add the biomechanical properties of each individual’s cornea (most importantly Young’s modulus, which is hard to measure in vivo), which are dependent on CCT, which is commonly studied in the clinic. Thinner corneas have inherently less resistance to applanation, making the relationship between thinner CCT and lower GAT measurements well known.[1,2,3,4,5,32] The present paper highlights how there is an additive effect of thinner corneas with higher altitudes, increasing the systematic measurement error of GAT, with greater potential underestimation of IOP, especially in glaucoma patients.

Most publications in this ecological meta-analysis are from cities with a low altitude. Using weighted means instead of simple means helps prevent the under-representation of the highest cities included and avoids over-representing eyes from very young patients, which may have happened with the few congenital glaucoma patients included.

An additional source of bias might result from different sequences of GAT versus DCT among the studies, number of observers, racial proportions, and situations impossible to control or analyze such as local changes in atmospheric pressures induced by climate and temperature. Measurements performed using DCT might be less affected by local atmospheric pressure changes because it sets to zero when it is turned on and thus measures gauge pressure in a similar manner as manometry. Another limitation that needs to be considered is that the formulas for corneal transplants and keratoconus are based on a smaller number of eyes as compared to the other groups, probably making them less precise in their estimates, thus warranting further clinical studies to better understand the relationship between GAT, CCT, and altitude in those eyes.

DCT measurements, as indicated by studies conducted by other research groups, exhibit a closer alignment with manometric (gauge pressure) invasive data.[16,17,18] Our findings underscore the critical need to account for environmental variables, such as altitude, when performing IOP measurements using GAT. This consideration becomes particularly pertinent for patients with glaucoma undergoing assessments in diverse geographic locations and for the interpretation of results in the context of multicenter investigations.

Finally, it is recommended that multicenter studies involving IOP measurements with GAT consider altitude above sea level as a variable or use the Pascal tonometer routinely. This will help to reduce the effect of environmental factors on IOP measurements and improve the accuracy of results in multicenter studies.

Conclusion

In summary, our study establishes that differences between Pascal and Goldmann IOP measurements are accentuated in cases of thinner corneas and at higher altitudes, with the notable exceptions of eyes post refractive surgery and corneal transplants. This outcome emphasizes the heightened risk of underestimating IOP not only in eyes with thinner corneas but also at elevated altitudes, particularly in individuals with glaucoma and those who have undergone corneal refractive surgery.

This comprehensive exploration of altitude’s influence on IOP measurements advances our understanding of the complex interplay between corneal biomechanics, atmospheric pressure, and tonometry methods. By shedding light on these intricate relationships, our study contributes to the refinement of IOP measurement protocols and strengthens the validity of research outcomes in clinical practice.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.