Assessment of the Radiopacity of Medication in Veterinary Medicine

ABSTRACT

Radiographic and computed tomographic (CT) visibility of human medications has been widely studied; however, equivalent data for veterinary medications remain limited. This study evaluated the radiographic visibility, dissolution characteristics, and CT features of commonly used veterinary and some human medications using a standardized water model. Seventy‐five medications were assessed for radiopacity prior to submersion, immediately after submersion, after 30 min, and after 1 h. All medications were visible without water. More than half (54.6%) remained identifiable immediately after submersion, decreasing to 13.3% after 30 min and to 6.6% after 1 h. The gastrointestinal complementary feed, vitamin B12, potassium chloride, and carprofen retained visibility throughout the experiment. Tablet area and height varied widely and demonstrated a weak positive correlation with visibility only after 60 min of submersion. Higher Hounsfield units (HU) values were moderately correlated with visibility at baseline, with the strength of this association decreasing over time as dissolution progressed. This study suggests radiography may be of limited use for the identification of ingested medication and provides images and HU values as a reference for clinicians.

1. Introduction

In human medicine, the radiopacity of numerous medications has been investigated using both radiography and computed tomography (CT) [1, 2, 3, 4, 5].

Various mnemonics, such as “CHIPS” (chloral hydrate, heavy metals, iron, phenothiazines, and slow‐release (enteric‐coated) preparations) and “BET A CHIP” (barium; enteric‐coated tablets; tricyclic antidepressants; antihistamines; chloral hydrate, cocaine, condoms, and calcium; heavy metals; iodides; and potassium–phenothiazines), have been proposed to aid clinical interpretation [1, 2]. However, these classifications are considered incomplete and inaccurate. Savitt et al. proposed a correlation with Hounsfield units (HU) and medication visibility—in which medications >1300 HU are consistently visible, 900–1300 HU variably visible, and <900 HU radiolucent [1].

The last major publication dates more than 20 years ago. The composition of several medications has changed since then, and there are various new medications on the market.

Accidental ingestion, incorrect administration, or uncertainty regarding medication intake are relatively common occurrences in veterinary practice. Anecdotally, radiography is used in veterinary emergencies to screen for the presence/absence of ingested medication. Few veterinary studies have investigated medication radiopacity [6, 7, 8]. To the authors’ knowledge, published reports include two case studies describing radiopaque tablets identified in canine stomachs—amoxicillin and bismuth subsalicylate—and one observational study examining the radiographic appearance of a small number of medications, including ampicillin, chloramphenicol capsules, diltiazem hydrochloride, aminophylline, dl‐methionine, liothyronine, strongid, and sucralfate. All of these have shown some degree of radiopacity. However, the range of medications tested was limited, and some formulations are no longer in use. For example, canine hypothyroidism is now typically treated with levothyroxine rather than liothyronine [9].

An improved understanding of the radiopacity and dissolution characteristics of current veterinary medications would assist clinicians in determining whether radiographs are useful when ingestion is suspected or when radiopaque foreign material with increased radiopacity in relation to its surroundings is noted in patients receiving oral medications.

The objectives of this prospective study were therefore to [1] evaluate the radiographic visibility of commonly used veterinary and selected human medications and determine the utility of radiography as a diagnostic modality for their detection; [2] determine how radiographic visibility changes over time during dissolution using a water model; [3] examine CT metrics, including HUs, area, and height, and their relationship to radiographic visibility. We hypothesized that most medications would initially have increased opacity in relation to the surroundings on plain radiography, but that the majority would lose visibility during dissolution, potentially limiting the utility of radiography for their detection.

2. Materials and Methods

A total of 75 medications were examined. Sixty‐four represented commonly used veterinary medications at Frontier Veterinary Specialists in Hergolding, Germany, and the remaining 11 drugs were human drugs found in the private household of two of the authors (E.D., A.M.A.). A detailed list of medications is provided in the Supporting Information. As this study did not involve live animals or animal‐derived data, ethical approval by the Committee on Ethics in Animal Use was not needed.

2.1. Experimental Setup

The experimental design was adapted from the method of Handy et al., whereby a container was filled with 20 cm of water, and the medication was submerged at a level of 10 cm [5]. In the present study, the medications were evenly spaced in a small sturdy radiolucent plastic container (container A) and secured to the container via a double‐sided adhesive tape. A second, larger plastic container (container B) was placed on the x‐ray table, and “container A,” with the secured medications inside, was placed inside “container B” at an elevation of 10 cm. “Container B” had a height of more than 20 cm to allow complete submersion of “container A,” which included the medications. Water was added subsequently into “container B” until a water column height of 20 cm was reached, ensuring complete submersion of “container A” and the secured medications.

2.2. Computed Tomography

For the first part of the study, the medications were scanned plain, that is, without water, to determine the HU values for each medication. Images were obtained using a 128‐slice multi‐slice CT scanner (Somaton Definition Edge, Siemens Healthineers, Forchheim, Germany).

The small container with the attached medications was placed on the CT table and scanned with the following settings: 120 kVp, 83 mAs, 1.0 mm slice thickness, 500 mm FOV, 94 mm reconstruction FOV, 512 × 512 matrix, and 0.6 pitch.

2.3. Radiography

A series of radiographs was obtained using the Cuattro DR HD system (Heska Corporation, Loveland, CO, USA).

The small container with the adhered medications was radiographed initially with no water [Tp]. The settings used were 80 kVp and 2 mAs.

The smaller container was then placed into the larger container at an elevation of 10 cm. The elevation was facilitated by the use of stacking material and tape. The large container was subsequently filled with water at a temperature of approximately 38°C to a total level of 20 cm.

Radiographs were acquired immediately after the container was filled [T0] using a kVp of 90 and mAs of 6. Radiographs were repeated after 30 min [T30] and 1 h [T60] with the same settings.

To avoid superimposition, the effervescent tablets were radiographed separately due to their rapid and excessive dissolution. Each effervescent tablet was placed in a glass cup containing 20 cm of water. For comparison, a glass of water without a tablet was placed beside.

2.4. Image Evaluation

The DICOMs were transferred to a workstation, and for interpretation, a standard DICOM viewer (Horos) was used. For the CT scans, the center of each medication along its long axis was visualized individually using MPR. Screenshots of the radiographs were made, and the images were labeled with alphabetic row labels and numeric column labels to improve identification of the medications’ positions.

2.4.1. CT Metrics

The edge of each medication in its long axis was outlined using a freehand region of interest (ROI) tool by one of the authors (second‐year resident, E.D.). The ROI was drawn with an approximately 1 mm margin inside the tablet boundary to avoid including surrounding air in the measurement. For smaller medications, where maintaining a full 1 mm margin was not feasible, the ROI was positioned as close as possible to the intended 1 mm boundary. This methodology was derived from a previously published study [10]. As tablets can contain an external coating, an additional ROI was placed along the peripheral edge of the medication. For both ROIs, the area (mm²) and the mean HU were recorded. In addition, the maximum vertical height (mm) of each medication was measured and documented.

2.4.2. Visibility of Medications

For each radiograph (0, 30, and 60 min) and medication, a subjective visibility score in relation to the surrounding water was determined by each observer, ranging from −2 to +2, as described in Table 1. The observers were given the labeled screenshots and the DICOMs for evaluation. The observers scored the images on their own initially. Where scores differed, these were then scored again by consensus at a later date.

TABLE 1.

Objective density scoring system used for visibility assessment.

Score	Opacity compared to the surrounding water
−2	Marked decreased opacity
−1	Mild decreased opacity
0	Not distinguishable
+1	Mild increased opacity
+2	Marked increased opacity

2.4.3. Statistical Analysis

Statistical analysis was performed using the R statistical software. Pearson correlation analysis was applied to evaluate the linear relationship between repeated measurements of HUs and area, obtained from the two distinct regions of interest. In addition, Spearman's rank correlation analysis was used to assess the association between radiographic visibility scores at different time points and quantitative parameters derived from computed tomography, including HUs, area, and size measurements.

3. Results

3.1. Medications

A total of 75 medications were examined, of which 64 were commonly used and available medications at the referral center, where the study was conducted, and the remaining 11 medications were human medications found in private households, which are displayed in row “7” (Figure 1). A variety of different active ingredients were included in this study. Some active ingredients were similar, but from different manufacturers. If available, medications with the same active ingredient and same manufacturer, but with a different concentration, were examined and placed side by side. A complete listing of the medications, including concentration and manufacturer, is found in the Supporting Information The majority of medications were tablets, seven were capsules, and one was a suppository.

FIGURE 1.

Photograph of the 75 examined medications within the smaller plastic container held in place by double‐sided adhesive glue. The position of the medications is constant throughout the radiographic study.

3.2. CT Results

3.2.1. Hounsfield Units

Using the 1 mm boundary method, HU values ranged from −586 to 3064 HU with a mean value of 249 HU. Measurements obtained with complete inclusion of the medication boundaries ranged from −585 to 2714 HU with a mean value of 269 HU.

HU measurements from the two regions of interest demonstrated an excellent positive correlation (r = 0.99, p < 0.001), indicating a high level of consistency between the measurement approaches.

In the following sections, HU measurements are reported first with the value obtained with a 1 mm internal margin, followed by the value obtained with the tablet edge included. The highest mean HU values for both measurements were observed in potassium iodide (7E, 3064 HU/2714 HU), followed by potassium chloride (3F, 1054 HU/1133 HU) and bethanechol chloride (4C, 1032 HU/1012 HU). Capsules exhibited the lowest HU values. The lowest values being vitamin B12 dietary supplement (7I, −586 HU/−585 HU); followed by Gabapentin (3C, −320 HU/−299 HU); and one combined preparation (paracetamol, ascorbic acid, caffeine, and chlorphenamin maelat) (7C, −306 HU/−296 HU). Potassium chloride (3F, 1054 HU/1133 HU) is the one exception. The remaining capsules (2G, 2H, 3E) had similar low HU values measuring −301 HU/−283 HU, −222 HU/−296 HU, and −100 HU/−195 HU, respectively. A complete listing of the HU values is found in Table 2.

TABLE 2.

Computed tomography‐derived mean Hounsfield units (HU) of 75 veterinary and human medications examined in this study, organized according to their position in Figures 1, 2, 3 (with the exception of the effervescent tablets radiographed separately in the water experiment).

	A	B	C	D	E	F	G	H	I	J	K
1	172/204	258/277	214/242	182/204	259/284	285/317	402/379	273/325	235/297	196/242	289/307
2	141/182	310/401	283/360	372/350	291/311	236/252	−317/−283	−344/−296	279/325	165/207	113/177
3	190/229	182/247	−320/−299	143/191	−222/−195	1054/1133	179/274	265/241	410/383	190/250	313/366
4	185/195	58/121	1032/1012	58/109	−100/−33	341/278	207/237	178/252	168/241	163/228	272/268
5	278/341	245/167	249/208	524/357	317/273	173/175	113/121	67/117	138/198	331/383	204/241
6	678/645	314/343	224/219	267/224	403/474	142/121	174/128	156/196	154/208
7	82/111	256/247	−308/−296	29/78	3064/2714	85/131	−33/−1	576/604	−586/−585	419/461	698/737

Note: Listed HU values of the medications visible throughout dissolution, with a positive visibility score (1K, 3F, 5J, and 5K), are highlighted in bold. HU value of the medication visible throughout dissolution, with a negative visibility score (2G), is shown in italics (2G).

Regarding the HU grouping previously implemented by Savitt et al., the majority of medications had values <900 HU (72/75; 96%), two had values between 900 and 1300 HU (potassium chloride and bethanechol chloride; 2.5%), and one medication was grouped at >1300 HU (potassium iodide; 1.5%).

3.2.2. Area

The area (mm²) in one centered transverse image of each medication was documented for both measurements. With inclusion of the 1 mm boundary, the measured areas ranged from 1.8 to 230.5 mm², with a mean value of 29.7 mm². The measured area with complete inclusion of the medications’ borders ranged from 7.3 to 287.6 mm², with a mean value of 52.7 mm².

Area measurements from the two regions of interest demonstrated an excellent positive correlation (r = 0.98, p < 0.001), indicating a high level of consistency between the measurement approaches.

In the following sections, area measurements are reported first with the value obtained using a 1 mm internal margin, followed by the value including the tablet edge. The medication with the largest area was the effervescent tablet in position 7H (230.5 mm²/287.6 mm²), followed by the gastrointestinal complementary feed (1K, 195.0 mm²/280.1 mm²). A complete listing of the measured areas is found in Table 3.

TABLE 3.

Computed tomography‐derived measured areas (mm²) of 75 veterinary and human medications examined in this study, organized according to their position in Figures 1, 2, 3 (with the exception of the effervescent tablets radiographed separately in the water experiment).

	A	B	C	D	E	F	G	H	I	J	K
1	19.9/34.5	107.0/161.3	20.1/45.2	33.3/59.6	34.2/60.4	7.0/16.1	4.5/13.9	23.4/42.3	4.2/15.6	12.1/30.0	195.0/280.1
2	56.7/89.9	4.6/27.4	18.1/45.9	2.0/14.4	3.3/18.0	13.0/28.6	21.5/46.3	37.7/68.2	7.4/23.6	4.5/18.2	19.5/35.2
3	5.5/18.2	16.5/43.4	59.1/96.9	57.1/99.1	37.7/89.5	52.2/102.6	2.6/13.8	5.9/20.9	3.3/15.5	43.3/75.1	21.0/42.7
4	11.3/34.4	28.5/50.8	15.0/32.6	16.7/35.7	30.2/61.4	5.0/21.9	2.9/15.2	6.5/20.2	14.5/32.7	27.9/50.8	7.7/22.2
5	5.8/18.5	3.3/15.4	2.5/14.7	1.8/7.3	7.0/21.7	3.2/11.3	5.6/18.1	8.2/23.6	14.3/41.8	22.5/42.2	96.0/136.9
6	2.7/11.0	7.5/21.0	21.1/43.8	13.8/28.3	14.1/40.2	6.5/21.1	7.0/25.9	28.2/57.6	45.3/88.5
7	53.3/87.0	7.4/21.1	70.9/110.7	18.9/41.9	6.8/24.3	16.6/42.0	104.9/159.7	230.5/287.6	45.6/65.5	93.7/133.6	82.1/123.3

Note: Listed areas of the medications visible throughout dissolution, with a positive visibility score (1K, 3F, 5J, and 5K), are highlighted in bold. Area of the medication visible throughout dissolution, with a negative visibility score (2G), is shown in italics (2G).

3.2.3. Height

The measured heights ranged from 1.3 to 9.2 mm, with a mean of 3.9 mm.

Similar to the CSA, the effervescent tablet (7H) demonstrated the largest height (9.2 mm), followed by the gastrointestinal complementary feed (8.3 mm). A complete listing of the height is found in Table 4.

TABLE 4.

Computed tomography‐derived measured vertical height (mm) of 75 veterinary and human medications examined in this study, organized according to their position in Figures 1, 2, 3 (with the exception of the effervescent tablets radiographed separately in the water experiment).

	A	B	C	D	E	F	G	H	I	J	K
1	3.0	6.6	3.8	4.4	3.6	3.2	2.1	3.2	1.3	2.5	8.3
2	4.5	2.3	3.7	1.6	2.3	3.4	4.1	4.2	2.0	2.6	3.8
3	2.8	4.7	5.8	7.2	6.5	6.7	2.6	2.9	2.4	5.3	4.9
4	3.0	4.4	3.6	3.9	4.9	2.2	2.6	3.8	4.2	4.6	2.5
5	2.9	2.5	2.3	1.6	2.5	2.6	3.6	3.7	3.2	4.1	8.2
6	2.3	3.7	4.9	3.3	3.1	2.7	2.7	4.2	5.6
7	5.4	2.8	6.0	4.2	3.5	4.0	7.5	9.2	4.6	4.6	4.8

Note: Listed heights of the medications visible throughout dissolution, with a positive visibility score (1K, 3F, 5J, and 5K), are highlighted in bold. Height of the medication visible throughout dissolution, with a negative visibility score (2G), is shown in italics (2G).

3.3. X‐Ray Results

3.3.1. Tp

All medications were visible in the plain radiographs, prior to submersion, and each was given a score of “+2” (75/75; 100%) (Figure 2A).

FIGURE 2.

Radiographs of the medication. (A) Plain radiograph, prior to water submersion. (B) Radiograph acquired immediately after submersion (0 min).

3.3.2. T0

Immediately after water submersion, more than half of the medications remained detectable (41/73; 54.6%), with the majority experiencing a reduction in the visibility score from “+2” to “+1” (31/41; 75.6%) (Figure 2B and 4A). The visibility remained unchanged for the gastrointestinal complementary feed (1K), potassium chloride (3F), carprofen 100 mg (5K), and potassium iodide (7E) (4/41; 9.8%). Gas opacity was present in a total of six medications, all being capsules (6/41; 14.6%). Vitamin B12, itraconazole, and vitamin B12—dietary supplement (2G, 3E, and 7I) reached a visibility score of “−2” and the remainder a visibility score of “−1” (2H, 3C, and 7C), corresponding to 3/6 (50%) for each group.

FIGURE 4.

Radiographs acquired after water submersion of the effervescent tablets. The tablets are placed in the left containers in each image. (A) Radiograph acquired immediately after submersion (0 min). (B) Radiograph acquired 30 min after submersion (30 min) and (C) radiograph acquired 1 h after submersion (1 h).

3.3.3. T30

Thirty minutes into the water model experiment, 10 of the previous 41 medications were still radiographically visible (10/75; 13.3%; Figure 3A and 4B). With the exception of vitamin B12 (2G) scoring “−1,” all medications were rated as “+1” and were tablets (1/10; 10% and 9/10; 90%, respectively). Potassium iodide was the only medication previously rated as “+2” that was now rated as “0” and not visible anymore.

FIGURE 3.

Radiographs of the medication. (A) Radiograph acquired 30 min after water submersion (30 min) and lastly, (B) is the radiograph acquired 1 h after submersion (1 h).

3.3.4. T60

After 1 h, 5 of the previous 10 medications were visible (6.6%; Figure 3B and 4C). Similar to the radiograph acquired 30 min after water submersion, vitamin B12 (2G) scored “−1” and the remainder “+1,” including the gastrointestinal complementary feed (1K), potassium chloride (3F), carprofen 20 mg (5J), and carprofen 100 mg (5K) (1/5; 20% and 4/5; 80%, respectively) (Figure 3B). Carprofen 20 mg was the only medication with a consistent visibility score throughout the study. These findings are highlighted in Table 5.

TABLE 5.

A complete list of the 75 veterinary and human medications examined in this study, organized according to their position in Figures 1, 2, 4 (with the exception of the effervescent tablets radiographed separately in the water experiment).

	1	2	3	4	5	6	7
A	Amlodipin 1.25 mg	Cefalexin 300 mg	Doxycyclin 50 mg	Metronidazol 250 mg	Pantoprazole 20 mg	Sotalolhydrochlorid 40 g	Butylscopolaminiumbromid 10 + Paracetamol 500 mg
B	Amoxicillin 400 mg + Clavulanic acid 100 mg	Cefalexin 250 mg	Enrofloxacin 150 mg	Mycophenolatmofetil 250 mg	Prednisolone 5 mg	Trimethoprim 20 + Sulfadiazin 100 mg	Colecalciferol 1000 I.E.
C	Amoxicillin 250 mg	Cefalexin 500 mg	Gabapentin 300 mg	Bethanecholchloride 25 mg	Prednisolone 50 mg	Trimethoprim 80 + Sulfadiazin 400 mg	Paracetamol 200 mg + Ascorbic acid 150 + Caffeine 25 + Chlorphenamin maelat 2.5 mg
D	Amoxicillin 200 mg + Clavulanic acid 50 mg	Maropitant 16 mg	Bentonit 1600 mg	Omeprazole 10 mg	Prednisolone 5 mg	Tramadol 80 mg	Ibuprofen 400 mg
E	Oclacitinib 16 mg	Maropitant 24 mg	Itraconazol 100 mg	Omeprazole 20 mg	Prednisolone 20 mg	Trazodon‐hydrochlorid 100 mg	Potassium iodide 65 mg
F	Acetylsalicylic acid 100 mg	Clindamycin 75 mg	Potassium chloride 600 mg	Ondansetron 4 mg	Pregabalin 25 mg	Torasemide 0.75 mg	Paracetamol 500 mg
G	Atenolol 25 mg	Vitamin B12 0.7 mg	Complementary feed	Robenacoxib 6 mg	Pregabalin 50 mg	Atenolol 25 mg	Paracetamol 250 mg
H	Vitamin B 240.5 mg	Phenoxybenzamine 10 mg	Marbofloxacin 20 mg	Robenacoxib 10 mg	Pregabalin 75 mg	Ursodesoxychol acid 150 mg	Effervescent tablet – Dietary supplement
I	Benazeprilhydrochlorid 5 mg	Dimazon 10 mg	Metoclopramidhydrochlorid 10 mg	Robenacoxib 20 mg	Spironolactone 50 mg	Pimobendan 5 mg	Vitamin B12
J	Benazeprilhydrochlorid 2.5 mg	Doxycyclin 50 mg	Meloxicam 2.5 mg	Robenacoxib 40 mg	Carprofen 20 mg		Effervescent tablet – Dietary supplement
K	Gastrointestinal‐ Complementary feed	Doxycyclin 200 mg	Metamizol 500 g	Phenybutazon 50 + Prednisolone 1.5 mg	Carprofen 100 mg		Effervescent tablet – Dietary supplement

Note: Medications highlighted in bold were visible throughout dissolution, with a positive visibility score (1K, 3F, 5J, and 5K). The medication shown in italics (2G) remained visible throughout the study but received a negative visibility score.

3.4. Correlation of CT Metrics With Visibility

3.4.1. Correlation of HUs With Visibility

On the basis of the classification by Savitt et al., potassium iodide was assigned to the consistently visible category (>1300 HU). It was visible immediately after submersion but became inconspicuous during dissolution. One of the two medications classified as variably visible (900–1300 HU), bethanechol chloride, remained visible for up to 30 min, whereas potassium chloride was consistently visible throughout the study. All remaining medications, classified as radiolucent (<900 HU), demonstrated variable visibility.

For consistently visible medications, HU values ranged from −317 HU/−283 HU to 1054 HU/1133 HU, with 5 out of 6 values below 900 HU. Medications visible for up to 30 min had HU values between 191 and 1011 HU, and most were also below 900 HU (4/5; 80%). Medications visible only immediately after submersion showed HU values from −586 HU/−585 HU to 3064 HU/2714 HU, again with the majority below 900 HU (30/31; 96,8%).

Higher HU values were associated with increased visibility, although this relationship was not consistent across all medications. Spearman analysis showed a moderate correlation for both measurements immediately after submersion (T0) (𝜌 = 0.44 and 𝜌 = 0.52, both p < 0.001), which decreased at later time points but remained statistically significant (T30: 𝜌 = 0.27 and 0.31; T60: 𝜌 = 0.27 and 0.28, all p < 0.05).

3.4.2. Correlation of Area With Visibility

Consistently visible medications had areas of 42.2–280.1 mm²; medications visible for up to 30 min, 11 to 99.1 mm²; and medications only visible immediately after submersion, 15.5 to 287.6 mm². Overall, the medication area showed no consistent relationship with visibility. This was reflected in the Spearman correlation analysis, which showed no significant association at T0 (𝜌 = 0.11 and 0.10, both p > 0.3) and at T30 (𝜌 = 0.16 and 0.15, both p > 0.1). At T60, weak correlations were observed (𝜌 = 0.23 and 0.22), reaching statistical significance only for the measurement with the ROI drawn around the medications’ periphery (p < 0.05).

3.4.3. Correlation of Height With Visibility

Consistently visible medications ranged from 4.1 to 8.3 mm in height, with three of five being among the tallest medications in this study. Medications visible for up to 30 min measured 2.3–7.2 mm, and those visible only immediately after submersion measured 2.3–9.2 mm. Overall, the taller tablets tended to be more visible, but this did not hold true across all tablets. This was reflected in the Spearman correlation analysis, which demonstrated no significant association at T0 (𝜌 = 0.13, p = 0.26) and at T30 (𝜌 = 0.21, p = 0.067), but a weak yet statistically significant correlation at T60 (𝜌 = 0.25, p = 0.029).

4. Discussion

The present study evaluated the radiographic visibility and radiopacity of 75 commonly used veterinary and human medications, both in their native state and during submersion in a water model. In line with previous studies, all medications were radiographically visible in undissolved form. Forty‐one of the 75 medications remained visible immediately after submersion, but visibility markedly decreased after 30 min and after 1 h. Only five medications were visible after 1 h submersion: gastrointestinal complementary feed, potassium chloride, carprofen (20 and 100 mg), and vitamin B12. These results support the hypothesis that most medications lose visibility during dissolution, thereby limiting the effectiveness of radiography as a diagnostic tool for identifying ingested medications.

Veterinary studies on the radiopacity of medications are limited and, to the authors’ knowledge, encompass only two case reports and one study. In one case report, a radiopaque bismuth subsalicylate tablet was identified within the stomach [8]. Bismuth is a heavy metal and radiopaque. In the other case report, an amoxicillin tablet—also located within the stomach—was identified on CT, measuring 470 HU; and 1000–1100 HU in an in vitro CT evaluation [7]. The amoxicillin preparations analyzed here ranged from 242.25 to 277.17 HU and were not readily visible radiographically. However, a direct comparison of the medications’ visibility is not possible, as the case report did not include radiographs. Regarding the difference in HU measurements, the authors suggest that these may result from variation in composition. The study by Hall reported radiopacity for several medications, including ampicillin, chloramphenicol capsules, diltiazem hydrochloride, aminophylline, dl‐methionine, liothyronine, strongid, and sucralfate [6].

Savitt et al. proposed that medications exceeding 1300 HU are consistently visible on radiographs, those between 900 and 1300 are variably visible, and those below 900 HU are radiolucent. In the current study, potassium iodide was the only medication exceeding 1300 HU, and initial visibility was high. This finding is in accordance with previous studies using iodide compounds. However, it was inconspicuous after 30 min and 1 h, likely due to its rapid solubility [11]. Two medications—potassium chloride and bethanechol chloride—fell between 900 and 1300 HU, both being detectable immediately after submersion and showing agreement with Savitt's classification. However, 72 of the 75 medications (96%) had an HU below 900 but exhibited variable radiographic visibility. For example, the gastrointestinal complementary feed, both concentrations of carprofen and vitamin B12, exhibited consistent visibility despite low HU values (−283 to 383 HU). This finding is contradictory to the classification. This discrepancy may be explained by the utilization of a digital system, as opposed to a film system, and the application of different exposure settings. The composition of medications is dependent on the manufacturer, and the authors suspect that different compositions can determine the radiographic visibility [12]. Seventy‐two of the 75 medications (96%) in this study were <900 HU. Despite this, they exhibited a variable visibility, suggesting that a classification based on HUs is not applicable. An overall tendency toward increased visibility with higher HU values was noted, but this was not consistent across all formulations.

This study used a simple relative opacity to score the medications using water as a baseline opacity and the medications as more or less opaque by comparison. This is similar to the visual grade analysis scoring method used when assessing new technology in medical imaging, and thus, the authors felt this was most appropriate [13, 14]. In addition, recent studies have shown that structures previously thought to be the same opacity as soft tissue structures may now be visible with the increased contrast resolution of digital radiography compared to film radiography [15]. This also influenced the decision to use relative scoring rather than absolute terms such as soft tissue opacity or mineral opacity.

Eight medications had HU values below zero: all capsules (except potassium iodide), omeprazole 20 mg, and paracetamol 250 mg. Capsules frequently exhibited negative HU values and air opacities on radiographs after submersion, consistent with the presence of air within the capsule shell. Paracetamol and omeprazole (20 mg) also showed negative HU values (−1 HU and −33 HU), and both are compact medications. In the radiographic experiment, they were not visible. The paracetamol in question was a suppository. Suppositories are frequently fat‐based, in this case with hard fat. As soon as the body temperature in the rectum melts the fat component, the active agent is released [16]. This presence of fat may account for the negative HU value. Likewise, fat‐like ingredients are found within the core of omeprazole, such as glycerol monoesterate, which could similarly contribute to the negative HU values observed in this medication. Interestingly, the lower dosed omeprazole formulation (10 mg, 4D) had positive HU values. This is explained by the smaller size of the lower dosed tablet, which results in an increased ratio of the nonfat‐containing coating to the fat‐containing core.

Regarding the capsules, the negative HU values are likely due to the air contained within, and this is in accordance with the negative values achieved in the water experiment.

Regarding the tablet size, a weak positive correlation was observed for larger tablets, but overall, medication size did not consistently predict radiographic visibility. For example, larger tablets, such as the effervescent preparations, initially appeared more radiopaque, but they dissolved rapidly, whereas smaller tablets like carprofen remained visible throughout the experiment. Interestingly, sotalol, one of the smaller medications in this study (area 2.7 mm²/11 mm², height 2.3 mm), remained visible for up to 30 min. Among the investigated medications, it was frequently observed that the size of the same medication from the same manufacturer increased proportionally to the active ingredients’ concentration. In most cases, the size did not influence the visibility. An exception was observed in the comparison of robenacoxib concentrations of 6 mg to the concentrations of 10, 20, and 40 mg. The investigation of the ingredients indicated that the latter three concentrations were accompanied by the presence of additional meat aroma and cellulose powder. Generally, the size and visibility were inconsistent, and therefore, similar to the measured HU, a classification based on the size of the medications was not applicable. It is, however, unclear whether size plays a factor in real‐life conditions.

To ensure that the CT‐derived parameters were not substantially influenced by ROI placement, two measurement approaches were evaluated. The measurements demonstrated strong positive correlations for both HU values and measured tablet areas, indicating that minor differences in ROI positioning did not affect the results. Inclusion of the tablet coating did not result in relevant differences in the obtained values. Overall, the observed relationships between CT‐derived parameters and radiographic visibility were similar for both measurements.

Regarding the composition of the inactive ingredients, multiple medications had similar compositions, and a correlation between the ingredients and radiographic visibility was not consistent. Manufacturers are, however, not required to include the amount of each inactive ingredient used [12]. Without this information, the comparison of radiographic visibility and dissolution is limited.

Mnemonics, such as “CHIPS” and “BET A CHIP,” have previously been proposed in human medicine [1, 2]. The formulas included partially overlap and are antihistamines, barium, cocaine, condoms, calcium chloral hydrate, heavy metals, iodides, iron, phenothiazines, slow‐release formulas (enteric‐coated preparations), and tricyclic antidepressants. Of these, only enteric‐coated preparations were included in the present study, and their radiopacity and visibility were found to be inconsistent. The inconsistency may be explained by advances in coating formulations, particularly since the original mnemonics date back to the late 1900s [17].

Radiographic detection of ingested medication is of particular importance in cases of suspected intoxication. According to the Veterinary Poisons Information Service (VPIS), the most recent published annual report (2022), NSAIDs, vitamin D, and vitamin C are among others, the most frequently reported causes of toxicosis in small animals [18]. Intoxications with NSAIDs, for example, ibuprofen, robenacoxib, carprofen, and meloxicam, are frequently reported in the literature, and multiple studies investigating the toxicity and treatment are available [19, 20]. In the present study, carprofen and meloxicam tablets retained radiopacity immediately after submersion, whereas ibuprofen and vitamin D rapidly lost visibility.

Effervescent tablets, with the exception of the tablet positioned at the location 7K (an effervescent calcium supplement), were multivitamin formulations and demonstrated a high initial visibility that diminished with dissolution, indicating that radiography could be valuable only shortly after ingestion. It is possible that the dissolution of the effervescent tablet could be slower in gastric juice, and visibility could be maintained over a longer period.

This study has several limitations. First, it is an in vitro study. The primary purpose of the water was to serve as a tissue equivalent phantom. The relative radiopacity of the medications to soft tissue was the primary goal of this study. Dissolution of medication was a secondary goal to better simulate the reality of emergency radiography—that is, an assumed delay between ingestion of medication and radiographic acquisition. The water method was selected as it was used in the limited literature on this topic [2, 4, 5].

An in vivo study would obviously be closer to the clinical picture, due to the variability in animal sizes and ethical concerns of using live animals (radiation, chemical restraint, etc.), the authors felt it best to perform an in vitro study from which results could be extrapolated. Intuitively, using hydrochloric acid (HCl) would appeal as an alternative to water. Although HCl is the main acid in gastric juices, the gastric juices are more complex than this, and thus, HCl is also not reflective of reality. HCl, unlike water, would also pose a risk to the authors and the equipment. Extrapolating from this study, it can be assumed that a medication that ceases to be visible in water after a period will also not be visible on radiographs in the stomach, where it is subject to mechanical and chemical digestion. Whether the persistently visible medication in vitro will also be visible in vivo is not definitive. Thus, this study should be viewed more as a list of medications that will not be visible on radiographs and those that may be visible on radiographs, and not a list of medications that will be visible on radiographs. The subset of possibly visible medications could be carried forward to a future study using live animals.

Second, all medications—with the exception of the effervescent tablets—were submerged in the same container and dissolved simultaneously, which could have caused cross‐contamination. The design was intentionally selected to ensure consistency and standardization of the experimental procedure throughout all medications, allowing for reliable comparison between samples. Notably, previous studies have not addressed this aspect, and a favorable outcome was assumed by the authors. Nevertheless, it cannot be excluded that cross‐contamination may have affected the results of this study.

Third, it is possible that the adhesive used to secure the tablets may have contributed to the superimposition of the radiographed medications. Furthermore, the presence of the glue could not be readily discerned from the medications in the CT scan, which suggests the possibility that it could partially be incorporated into the measurements.

Fourth, small air bubbles occasionally overlapped with the medications, possibly influencing visibility scoring, especially if they superimposed the frequently air‐dense capsules, leading to a false positive score. In order to mitigate the impact, a visual aid was provided to the observers, comprising an image containing numbers and letters, with the intention of facilitating the identification of the location of the medications.

Finally, a limited number of medications and of specific manufacturers were used. The medications included in this study were selected based on their availability in our clinic, with the VPIS annual report on commonly reported medications causing intoxications taken into consideration [18].

In conclusion, this study provides comprehensive reference data on the radiodensity and dissolution characteristics of medications commonly used in veterinary medicine. The low number of visible medications after 1 h of submersion in water calls into question the value of radiography in the majority of accidental ingestions. The results and reference images provided can aid practitioners in determining if radiographs are beneficial in cases where there is a suspicion of falsely ingested medications or medication‐induced intoxication. Readily visible medications were the gastrointestinal complementary feed, vitamin B12, potassium chloride, and carprofen.

Author Contributions

Conception and design: Eleonora D'Incecco and Conor Rowan. Acquisition of data: Eleonora D'Incecco. Analysis and interpretation of data: Eleonora D'Incecco, Anna M. Adrian, Conor Rowan. Drafting the article: Eleonora D'Incecco, Anna M. Adrian, Conor Rowan. Revising article for intellectual content: Eleonora D'Incecco, Anna M. Adrian, Conor Rowan. Final approval of the completed article: Eleonora D'Incecco, Anna M. Adrian, Conor Rowan. Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved: Eleonora D'Incecco, Anna M. Adrian, Conor Rowan.

Disclosure

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information: vru70183‐sup‐0001‐SuppMat.docx

Data Availability Statement

Data supporting the results are available from the corresponding author upon reasonable request.