Trends in Estimated Thyroid, Salivary Gland, Brain, and Eye Lens Doses From Intraoral Dental Radiography Over Seven Decades (1940 TO 2009).

The purpose of this study is to support retrospective dose estimation for epidemiological studies by providing estimates of historical absorbed organ doses to the brain, lens of the eye, salivary glands, and thyroid from intraoral dental radiographic examinations performed from 1940 to 2009. We simulated organ doses to an adult over 10 y time periods from 1940 to 2009, based on commonly used sets of x-ray machine settings collected from the literature. Simulations to estimate organ dose were performed using personal computer x-ray Monte Carlo software. Overall, organ doses were less than 1 mGy for a single intraoral radiograph for all decades. From 1940 to 2009, doses to the brain, eye lens, salivary glands, and thyroid decreased by 86, 96, 95, and 89%, respectively. Of these four organs, the salivary glands received the highest doses, with values decreasing from about 0.23 mGy in the 1940s to 0.025 mGy in the 2000s for a single intraoral radiograph. Based on simulations using collected historical data on x-ray technical parameters, improvements in technology and optimization of the technical settings used to perform intraoral dental radiography have resulted in a decrease in absorbed dose to the brain, eye lens, salivary glands, and thyroid over the period from 1940 to 2009.

INTRODUCTION

Ionizing radiation (IR) plays an essential role in medicine and dentistry by improving diagnosis and supporting patient care. The use of x rays as a diagnostic tool in almost all fields of medicine has resulted in an increase in IR exposure for the general population (Hall and Brenner 2008). Medical radiation has become the main source of man-made IR exposure (UNSCEAR 2000a and b). IR is known to cause several types of detrimental health effects, including carcinogenesis (NRC 2006). Thus, this increase in the use of IR in medicine has raised public health and radiological protection concerns.

The use of IR in dentistry began in 1896 (Todd 2014) and had become common practice by the 1950s due to its powerful diagnostic capabilities. Imaging of the teeth and supporting bone permits detection of many clinical conditions (e.g., caries, gingival and osseous diseases) (Iannucci and Howerton 2006). Dental x rays have become common practice and a routine part of patient evaluation in dentistry. The common use of x-ray imaging in dentistry is evidenced by the large numbers of images obtained as well as by the number of radiological procedures per patient (Wrzesien and Olszewski 2017; UNSCEAR 2008): in 1970 in the United States, 59 million people underwent a total of 278 million dental radiographs, corresponding to approximately 5 radiographs per patient (Lee 1974). The number of dental radiographs is increasing worldwide: in 1988, 340 million dental radiology examinations were performed, while in 2008 this number increased to about 480 million (UNSCEAR 2008).

Efforts worldwide have been made to keep the radiation dose of each radiograph as low as possible with the goal of minimizing patient exposure while obtaining an image of adequate quality. Technological improvements and optimization of x-ray technical and operating parameters throughout the last century have improved image quality while reducing exposure (Doi 2006). However, while the relationship between radiation and carcinogenesis at low doses and low dose rates remains uncertain and difficult to quantify, exposure to low-dose levels of IR could potentially result in adverse stochastic effects (NRC 2006). Organs that may present a particular concern in dental radiography are the brain, the lens of the eye, the salivary glands, and the thyroid because they are near the primary beam, and they are known to be radiosensitive. Multiple or frequent exposures to dental x rays have been found to be associated with an increased risk of thyroid cancer (Memon et al. 2010), brain cancer (Picano et al. 2012; Lin et al. 2013; Claus et al. 2012; Preston-Martin and White 1990), and salivary gland tumors (Preston-Martin and White 1990; Horn-Ross et al. 1997; Preston-Martin et al. 1988).

Risk estimations in the low-dose range are challenging; large epidemiological studies are required to reach an adequate study power (NRC 2006). Some published risk estimates rely on the total number or frequency of dental radiographs, a proxy for IR exposure. However, the radiation dose quantity that is most appropriate for cancer risk estimation is organ absorbed dose, defined as the mean energy imparted to the organ mass by ionizing radiation (NRC 2006). Absorbed dose is measured in gray (Gy), which is equivalent to J kg−1. Estimating dose on an individual basis is extremely challenging because neither the delivered doses nor the technical parameters used are routinely recorded in medical databases, particularly for dental x rays. When subject information comes from questionnaire data or medical records with no information on radiation dose, an alternative that can be used to support risk estimates in epidemiological studies is a dose estimate based on typical technical parameters (i.e., machine settings) in specific time periods (Chang et al. 2017; Melo et al. 2016; Preston-Martin and Pogoda 2003; Mathews et al. 2013).

In this study, we estimated the organ dose received by the brain, lens of the eye, salivary glands, and thyroid during adult intraoral dental radiography using technical parameters derived from the literature that we believe were commonly employed between 1940 and 2009. Dose estimation was performed with the Personal Computer X-ray Monte Carlo (PCXMC 2.0) software (Radiation and Nuclear Safety Authority [STUK], Helsinki, Finland) (Tapiovaara et al. 1997) using the Excel spreadsheet application (Microsoft Corp., Redmond, Washington, US).

MATERIALS AND METHODS

Definition of dental procedures of interest in this study

In dental intraoral radiography, the x-ray film (or in modern practice, the digital imaging sensor) is placed inside the patient’s mouth to obtain an image of the surrounding teeth, gum, and bone (Iannucci and Howerton 2006). A complete mouth examination is performed by obtaining multiple intraoral radiographs until the entire dentition has been imaged. In this study, we simulated only absorbed doses from intraoral radiography as our focus was on the years before cone-beam computed tomography (CT) and other techniques became commonplace.

Sources of information

We sought national and regional reports of radiological practice, radiological textbooks, and studies describing common radiological techniques in a particular decade. We conducted a literature search in PubMed, entering the following keywords: radiology, radiography, dental, x ray, organ dose. Additional publications were found by reviewing the reference lists of publications of interest. Data from dental radiology textbooks and guidebooks were also collected as available at our university medical library (Universitat de Barcelona). A set of inclusion criteria was defined, as described in Table 1, to select appropriate studies and reports. We evaluated the representativeness of the reported technical parameters using a relevance score. A high relevance score was given to studies based on nationwide collection of parameters commonly used in dental practice. A medium relevance score was given to (1) textbooks and guidebooks for undergraduate students and dentists in practice, and (2) studies describing practice in a region or a limited number of facilities. Publications reporting ideal settings were given a low relevance. We also gave higher priority to publications that provided a complete set of machine settings, to avoid unrealistic combinations of technical parameters.

Table 1

Inclusion criteria for reports/studies selection.

Data extraction

The following information was collected from the selected studies: the full reference, country and decade of examination, information about the study design, and the technical parameters needed for the organ dose simulation (as listed below). If, for a given decade, we could not find any publication that provided a complete set of technical parameters, we gave priority to the most complete ones by using additional publications from the same decade to complete the set. For instance, for the 1940s decade, values were extracted from two publications (Clark 1949; Sante 1949). Values provided in each publication were considered as one set and were completed with filtration and beam area values provided by Jamieson (1952) and Richards (1958), respectively. When a specific parameter could not be found for a given decade, we used the value given in publications for the previous decade.

Input parameters

The dose absorbed by a specific organ is estimated by the software using the following inputs: (a) factors based on patient characteristics such as age and body mass; and (b) a series of x-ray machines settings defined by the practitioner, including tube potential (x-ray voltage), filtration, focus-to-skin distance, projection angle, oblique angle, x-ray beam height and width, and the exposure.

Tube potential is measured as peak kilovoltage (kVp) and represents the potential applied across the x-ray tube. Filtration is measured in millimeters of aluminum (mm Al) and corresponds to the equivalent thickness of aluminum placed in front of the x-ray tube to filter out low-energy x-ray photons. The focus-to-skin distance (cm) is the distance between the x-ray source and the patient’s skin. X-ray beam height and width correspond to the size of the x-ray beam as it exits the collimator. The exposure corresponds to the radiation dose at the entrance surface of the patient. PCMXC software allows entry of this variable in any of five different units (mGy free in air, mR free in air, mGy cm−2, rad cm−2, mAs). In this study, we entered the tube current time product (milliamp seconds, mAs), hereafter referred as tube current.

Values of these parameters have changed over time due to continuing technological advances. To capture these changes, the technical parameters listed above were collected for each decade from 1940 to 2009.

Dose estimation

PCXMC software was used to simulate brain, salivary gland, and thyroid doses, using as input the technical parameters considered to be most commonly used in each decade between 1940 and 2009. The software calculates absorbed dose to 29 different organs and is well suited to simulate a wide range of radiological procedures including intraoral dental radiography (Lee et al. 2016; Aps and Scott 2014; Koivisto et al. 2012; Lindfors et al. 2017; Vassileva and Stoyanov 2010). PCXMC does not provide dose estimates for the lens of the eye. However, this organ dose value can be approximated by the entrance dose to the skin at the level of the eyes, as confirmed by the estimated tissue dose per unit air kerma (DT/Ka) factors for the lens of the eye (Simon 2011), for diagnostic x-ray energies, which are centered around unity. Eye lens doses were estimated by deriving the air kerma from the PCXMC-calculated brain dose using the conversion coefficient between brain dose and air kerma provided in Table 7 of the same publication (Simon 2011) for 70 kV as the most relevant tube potential for dental procedures. The air kerma at the level of the eyes was therefore considered a reasonable estimation of the eye lens dose.

Once sets of typical parameters for each dental projection were collected for each time period, simulations with PCXMC 2.0 were performed. When possible, for each decade we performed a simulation by cone (collimator) type (short or long), for each projection (maxillary or mandibular teeth), and for each tooth (incisor, canine, molar, or premolar). Therefore, for most decades we performed at least 16 different simulations. Median values were reported and considered as representative for each decade. Patient sex and age were set at male and 30 y (PCXMC uses an adult model, with a reference body mass and height for a male of 73.2 kg and 178.6 cm). The position of the phantom towards which the central beam was directed was set at the level of the mouth, using the software reference points Xref = 0, Yref = −5, and Zref = 80. Table 2 summarizes the applied projection and oblique angles for each target tooth (Williamson 2006).

Table 2

Angles used in simulations of intraoral examinations for the visualization of different teeth.

The use of patient-shielding devices was not considered in this study though it is widely understood that lead thyroid collars can reduce the dose to the thyroid gland substantially. This is especially true when anterior maxillary teeth are exposed. In those cases, thyroid collar use reduces thyroid dose levels around 75% (Hoogeveen et al. 2016). The extent to which protective devices have been used over time in different facilities is unknown. Therefore, for the dose calculation we assumed no shielding was used.

RESULTS

Of the 55 publications reviewed, we selected 18 that fulfilled the inclusion criteria. Table 3 summarizes the characteristics of the selected studies. For each decade, the number of relevant references used was two or three. Among the selected studies, only one study was scored with low relevance (Richards 1958), the majority were judged to be of medium relevance, and seven studies were scored as highly relevant. The high-relevance studies were generally descriptions of more recent practice. The technical parameters used to perform the simulations are reported in Table 4 for each of the seven decades between 1940 and 2009. Median organ dose estimates for the brain, eye lens, salivary glands, and thyroid for a single intraoral radiograph are provided together with ranges in Table 5. The ranges provided in Table 5 reflect the variability of doses which we have estimated from all combinations of parameters. The wide range of doses in the first decade reflects the very different tube current values reported for the period in the two available reports (Clark 1949; Sante 1949).

Table 3

Summary of the characteristics of the publications from which the technical parameters were obtained.

Table 4

Technical parameters used for organ absorbed dose estimation. References are provided as author and date.

Table 5

Derived organs absorbed doses (mGy) to the brain, eye lens, salivary glands, and thyroid.

The salivary glands received the highest absorbed doses, ranging from about 0.23 mGy in the earliest decade evaluated (1940–1949) to about 0.025 mGy in the most recent decade evaluated (2000–2009). Estimated absorbed doses to the brain were the lowest among the studied organs, ranging from 0.014 mGy in the 1950s to 0.00088 mGy in the 2000s.

The percentage change in dose from the previous decade was calculated and is reported in Table 5. For the four organs studied, the largest decrease in dose was seen between the 1990s and the 2000s (around 80%). Table 6 provides comparisons of our results to organ doses reported in previous publications.

Table 6

Absorbed organ dose value comparisons with previously published articles. Values in bold are comparable values.

Figs. 1 to 4 show the absorbed dose to the selected organs by decade. Each point represents the dose received by the organ of interest from an intraoral radiograph performed to visualize a specific tooth (incisor, canine, premolar, molar) from each projection (maxillary and mandibular) using both techniques (short and long cone). The dose range represents the variability obtained from simulating the procedure (as described above) with the set of parameters relevant for the given decade. For all organs studied, there was a trend towards decreasing dose over time.

Fig. 1

Absorbed doses (median and interquartile range) to the brain due to single intraoral examinations between 1940 and 2009.

Fig. 4

Absorbed doses to the thyroid (median and interquartile range) due to single intraoral examinations between 1940 and 2009.

Absorbed doses to the lens of the eye (median and interquartile range) due to single intraoral examinations between 1940 and 2009.

Absorbed doses to the salivary glands (median and interquartile range) due to single intraoral examinations between 1940 and 2009.

DISCUSSION

The absorbed doses reported in our study characterize the general trends in doses received from intraoral dental radiography as they evolved over time. Our methodology was designed to provide values of doses representative of exposure of the general adult population within each decade. Overall, the dose from intraoral radiography was very low; none of the median organ doses estimated here exceeded 1 mGy. However, intraoral radiography is performed often throughout a patient’s lifetime. For instance, the American Dental Association recommends regular routine intraoral examinations with bilateral posterior bitewing radiographs every 12 to 24 mo for children, every 18 to 36 mo for adolescents, and every 24 to 36 mo for adults who are recall patients with no clinical caries and no increased risk of caries (ADA/FDA 2012). Based on these recommendations and our estimates of organ dose, this means that, on average, a 70-y-old male in 2010 with no history of caries and no increased risk for caries development could have received 20 dental radiographic examinations during adulthood, each with bilateral posterior bitewing radiographs, leading to brain, eye lens, salivary gland, and thyroid cumulative absorbed doses of approximately 0.5 mGy, 2 mGy, 13 mGy, and 6 mGy, respectively. A similar adult with a history of caries or increased risk for caries development could undergo dental radiography with bilateral posterior bitewing radiographs at intervals of 6 to 18 mo according to current American Dental Association (ADA) recommendations, with resultant greater absorbed doses to the brain, eye lens, salivary glands, and thyroid (ADA/FDA 2012). Of course, this estimation assumes that the recommendations and practices with regard to the frequency of dental radiography were the same in previous decades.

The dose values presented in this report are for a single radiograph. As noted above, routine dental radiography typically includes bilateral posterior bitewing radiographs. In some clinical situations, it is necessary to obtain a picture of the entire dentition, leading to the performance of full-mouth examinations. Throughout the studied decades, the number of images necessary to obtain a full-mouth examination has been reported as between 14 and 27 (Alcox and Jameson 1974; Richards and Webber 1964; Underhill et al. 1988; Gofman and O’Connor 1985; Lee 1974; Weissman and Sobkowski 1970; Stanford and Vance 1955; Baily 1957; Lecomber and Faulkner 1993). This number of images would entail a concomitant increase in absorbed organ dose.

The decision to evaluate brain, eye lens, salivary glands, and thyroid dose was based on the biological relevance, radiosensitivity, and proximity to the radiation beam of these organs. Another organ of interest in radiation epidemiology is active bone marrow, also known for its radiosensitivity. However, we are not presenting any results on this organ because the amount of bone marrow at the dental arch is negligible (Cristy 1981), and as expected, PCMCX estimated doses for the bone marrow were close to 0 (data not shown).

Trends over time

In every decade studied, the salivary glands received the highest absorbed dose among the four organs studied. Overall, there was a decreasing trend over time in the dose received by each of the organs evaluated. This can be explained largely by technological changes in x-ray equipment and imaging techniques. Between 1940 and 2009 these changes resulted in a decrease of about 90% in organ dose from intraoral imaging. One important factor for this decline was the increase in total filtration to at least 2.5 mm Al for tube potentials higher than 70 kV, first implemented between the 1970s and 1980s. This decrease in exposure is related to changes in regulations (Iannucci and Howerton 2006; Whaites 2006; Mason 1988; Browne et al. 1995), based on the guidelines provided by the International Commission on Radiological Protection (ICRP) and in the United States, the National Council on Radiation Protection and Measurements (NCRP). Those two agencies represent, since the 1930s, the main advisory organizations for the limits for ionizing radiation exposure and protection of the public (IoM 1996). Reduction in the tube current (mAs) was a result of the introduction of faster films. Faster films require less radiation than slower films to produce the same film blackening. For instance, Kodak films available in 1980 reduced patient exposure over 95% compared with those available in 1940 from the same manufacturer (Richards and Colquitt 1981). Moreover, the subsequent introduction of F-speed films allowed reduction in patient exposure by half compared with E-speed films without loss of image quality (Farman and Farman 2000). Most recently, the advent of digital receptor technology has provided the opportunity for even further dose reduction (Farman and Farman 2005).

The estimated dose in the 1940s was lower than the dose in the decades that immediately followed. This might be explained by the lack of availability of data for the 1940s as only two sources of data for this decade (both of medium relevance) could be identified. Therefore, the dose estimates for this decade should be used with caution.

Comparison with previous studies

We found several publications that reported absorbed organ doses due to intraoral examinations. These publications allowed comparison with our results, as shown in Table 6. There were clear methodological differences, including the choice of technical parameters, phantom material, and dosimeter type and positioning compared to our calculations, and these could explain some of the differences with our results. When these publications provided doses only for specific projections, comparison was made to the corresponding estimated values.

The doses provided by Ekestubbe et al. (2004), for a comparison of absorbed and effective doses between scanographic and zonographic examinations with intraoral periapical radiography, are higher than what would be expected for the 2000–2009 decade. Their results are explained by their use of a higher tube current (15 mAs and 9.4 mAs for the maxillary and mandibular molars, respectively).

For the decade 1960–1969, the organ doses provided by Richards and Webber (1964), using phantom measurements in the head and the neck, are much lower than our estimates. As mentioned by Ekestubbe et al. (2004), the considerable differences in exposure due to changes of the x-ray beam direction relative to the position of the ionization chamber could explain this difference. Furthermore, they report that their results were lower than the ones that had been presented previously in the dental literature.

Sources of uncertainty

In dental radiographic procedures, the thyroid gland, brain, and eyes are near the edge of the radiation field, so doses to these organs can be affected by patient positioning. Slight positioning differences can result in important changes in absorbed dose to these organs (Kapila 2014). The difference between an organ falling completely inside the beam or on its edge can produce substantial variation in the resulting dose.

We present absorbed dose values for a single intraoral radiographic projection. However, in routine practice more than one projection may be obtained, depending on the clinical indication. The number of images recommended to obtain a complete mouth series varies, even within an individual decade. This variation is especially important for epidemiological studies where retrospective dose estimations are performed and cumulative doses are required. The dose values for a complete mouth series are derived from multiplication of the dose for a single intraoral radiograph by the number of radiographs obtained. Also, retakes of the image within the same clinical examination may be performed if the initial image is judged to be inadequate. Information on frequency of examinations may be collected through self-administered questionnaires or medical records reviews, but in both cases the exact number of projections may not be reported precisely. This can introduce additional uncertainty.

Another factor that influences doses to nearby organs is collimation. Both the degree of collimation and the beam shape (circular or rectangular) are important determinants of these doses. Although some of the sources we consulted specified the type of collimation used (circular or rectangular), we were restricted to defining beam height and width in a rectangular field in PCXMC. We have converted the diameter of circular beams to height and width values for a rectangle, making our estimations more conservative.

Kilovoltage values used for our simulation are those that dental radiographers were reported to have used or that were recommended, but values delivered in practice might have differed. In some detailed surveys, both the selected and the measured tube potential were provided. Selected values are those that the technician sets (theoretical ones); measured values are those actually generated by the radiographic system. Although measured tube potential is more accurate for estimating absorbed dose, we used the selected values for our simulations to allow better comparison among published studies. The 1999 Nationwide Evaluation of X-ray Trends (NEXT) Survey in the United States determined a mean absolute value difference between measured and selected values of intraoral tube potential of 3.8 kV (Moyal 2003).

Limitations

The major limitation of the present study was the paucity of relevant information in the literature for some of the time periods. Most articles that reported estimates of dose compared the impact of using several technical parameters with the goal of optimizing settings to reduce patient dose. Very few publications provided representative values that corresponded to common practice in a specific decade. The doses estimated in this work were based on typical technical parameters provided by a series of scientific publications, reports, and guidebooks for the last seven decades and a series of assumptions made to reconstruct typical protocols when only partial information was provided in the literature. The values we used for the various parameters are believed to be representative of those used in a high health-level country (UNSCEAR 2000), but variation around these values is expected between different countries, facilities, and radiographers. We considered the most common projections used to perform intraoral radiography, but alternative examination settings could also have been used, and this could have resulted in a modest variation in dose. For example, the difference in the dose medians between two sets of examination settings, both reported as common during the 1990s (Moyal 2003), was 0.0017 mGy, 0.0066 mGy, 0.049 mGy, and 0.013 mGy for the brain, eye lens, salivary glands, and thyroid, respectively.

Additional limitations related to data sources include our restriction to publications in the English language and our exclusion of reports from countries other than United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) level I. This means that our dose estimates may not be applicable worldwide or in less developed countries.

We calculated our estimates of organ dose using the PCXMC adult model with a reference male. We have not estimated absorbed dose to the relevant organs for children or adolescents because of the very limited data on typical technical parameters for dental radiography for those age groups. Doses that we report for adults should be used in epidemiological investigations of children with great caution.

We present estimates for median absorbed dose to each organ in each decade. As is apparent from Figs. 1–4, there was likely substantial variation in absorbed dose depending on which tooth was radiographed, especially in earlier decades (supplemental digital content [SDC] Appendix SDC 1, http://links.lww.com/HP/A172). Use of the median absorbed dose may result in systematic errors when only specific portions of the dentition are examined, such as in current American Dental Association recommendations (ADA/FDA 2012) where routine radiography consisting of bilateral posterior bitewing radiographs is recommended. Recommendations or common practice regarding routine examinations might have differed throughout the studied decades regarding the portion of the dentition examined.

CONCLUSION

Our estimates of absorbed organ dose may be used in epidemiological research either where the study seeks to examine the association between the radiation delivered during dental diagnostic procedures and the risk for development of cancer or lens opacity or where dental irradiation must be accounted for as part of the total exposure profile.

Dental procedures, as shown in the present work, have generally delivered very low doses, especially in recent decades, but their extensive and repeated use in the general population raises public health and radiological protection concerns. By collecting information on the frequency of dental examinations, epidemiological studies may estimate the health effect of such low-dose exposures. Studies that aim to predict lens opacity or cancer incidence attributable to diagnostic x-ray examinations may benefit from the results presented here. Such studies would require collecting, for each decade, detailed information on the frequency of each type of dental examination separately (intraoral periapical or bitewing, cephalometric, and panoramic x ray).