What Should the Optimal Adrenaline Auto-Injector Needle Length Be?

Objective: To identify the possible subcutaneous and peri/intraosseous injection risks with different needle lengths in adults using adrenaline autoinjectors (AAI) with different needle lengths. In addition, optimum needle lengths with the lowest injection failure risk were determined for both genders. Materials and Methods: Skin-to-muscle distance (STMD) and skin-to-bone distance (STBD) were measured under minimal and maximum pressures at the anterolateral aspect of the thigh by ultrasound. Risks of subcutaneous injection, peri/intraosseous injection, and total events were calculated for both genders by taking a 15.2 mm needle length as reference for calculations. Possible subcutaneous injection and peri/intraosseous injection risks with different needle lengths were calculated and the most optimal needle lengths with the least total event risk were determined for each gender using different body mass index (BMI) threshold levels. Results: A total of 208 adults (118 women, 90 men) with a mean age of 46.3±15.7 were enrolled. BMI had the highest correlation with all ultrasonographic measures for each gender. STMDmax of 45 women (38%) and one man (1.1%) were longer than 15.2 mm. For men, total event risk was similar (3.3%) with needle lengths ranging from 12 to 18 mm. For women, considering two different BMI threshold levels with two different needle lengths (20 mm for 28 kg/m2 and 26 mm for 32 kg/m2) reduced total event risk to 6.7% for both BMI thresholds. Conclusion: The risk for inadequate intramuscular delivery is higher in women, especially those with higher BMI. Individualization of AAI needle lengths according to certain BMI values can improve the outcome, especially for obese women.


INTRODUCTION
Despite the advances in understanding the underlying mechanisms of allergic diseases and the dazzling advances in the treatment of these diseases, the first drug of choice has never changed for one of the most severe allergic reactions. Adrenaline (epinephrine) is the initial drug of choice for anaphylaxis and delayed administration of adrenaline has been associated with increased risk of death (1)(2)(3). Nevertheless, it has been shown that only 67% of cases of near-fatal anaphylaxis received prompt adrenaline. Besides prompt administration, an appropriate route of administration is also important for rapid adrenaline efficacy. Intramuscular (IM) injection of adrenaline into the vastus lateralis muscle has been shown to result in quicker time to peak plasma concentration (Tmax) and higher peak plasma concentrations (Cmax) when compared to intramuscular or subcutaneous injection into the upper arm (4,5). Thus, The European Medicine Agency recommends an intramuscular injection of adrenaline as the preferred route of administration in the case of anaphylaxis (6).
The administration of adrenaline before admission to the emergency room has been shown to reduce the likelihood of hospital admission when compared to the administration at the emergency room (7). Thus, adrenaline auto-injectors (AAIs) which enable selfadministration outside the healthcare settings, have been developed (7). Different manufacturers have designed AAIs that basically differ in the drug dosage, trigger pressure and needle length. There are studies that show the needle length of these devices may be inadequate for IM injection, particularly in women and obese patients (8)(9)(10)(11). It was shown that women were 6.4 times more likely than men to receive adrenaline in the subcutaneous tissue (11). In their study, Bhalla et al. showed that those patients with increased subcutaneous injection risk had higher body mass index (BMI) (11). Similarly, in a significant number of pediatric patients, it was shown that the needles of AAIs might not be long enough to reach the intramuscular area (12). Because of the significant effect of gender and anthropometric characteristics on adequate IM administration, the need for age, gender and/or BMI specific auto-injectors was reported (13).
Prompt and effective intervention is crucial in acute and life-threatening reactions such as anaphylaxis, which mostly occurs outside the healthcare settings. AAIs should be able to deliver the whole drug dose intramuscularly to all patients of different genders and anthropometric characteristics as rapidly as possible and without hesitation. In the present study, we aimed to identify the possible subcutaneous and peri/intraosseous injection risks with different needle lengths, taking a 15.2 mm needle length as the reference for calculations also, based on the ultrasonographic measurements in adult patients. In addition, optimum needle lengths with the lowest injection failure risk were determined for both genders based on the results.

Patients
This study was carried out in the two largest health facilities of Kayseri, Turkey (Erciyes University Hospital and Kayseri City Hospital) between 2017-2018. Patients older than 18 years of age, who were admitted to the radiology clinic for ultrasound/doppler ultrasonography and who agreed to attend the study were included. Written informed consent was obtained from each participant. Patients with skin problems at the measurement area (pretibial edema, lymphedema or previous surgery of the leg) or with any neuromuscular problems that might cause muscle atrophy were excluded from the study. Age, gender, height, weight, and BMI of the patients were collected. The study was approved by the Erciyes University Ethical Committee, Kayseri, Turkey (Decision No: 2017/560).

Ultrasonographic Measures -Skin-to-Muscle Distance (STMD) and Skin-to-Bone Distance (STBD)
All measurements were done at the anterolateral aspect of the mid-right thigh using the Aplio-500 ultrasound system (Toshiba Medical Systems, Otawara, Japan) with 7,5 MHz linear probe by two radiologists. Both radiologists had similar measurements in 30 randomly selected patients. STMD and STBD were measured under minimal and maximum pressures. Minimal pressure ( min ) was defined as "just enough" pressure to obtain an adequate ultrasound image, and maximum pressure ( max ) was defined as the maximum pressure that can be performed by the radiologist and tolerated by the patient. The STMD was measured from the skin surface to the outer border of the vastus lateralis muscle, including the muscle fascia. The STBD was measured from the skin surface to the outer aspect of the periosteum of the femur.

Data Analysis
Risks of subcutaneous injection due to inadequate needle length, peri/intraosseous injection due to excessive needle length, and total events (sum of subcutaneous injection risk and peri/intraosseous injection risk) were calculated for both genders. Fifteen point two mm needle length of Epipen®, which is used worldwide, was taken as the reference for calculations. In addition, possible subcutaneous injection and peri/intraosseous injection risks with different needle lengths ranging from 12 to 28 mm were calculated in the same patient group. Finally, the anthropometric measurement that most correlated with the ultrasonographic measures was identified and total event risks were calculated for different threshold levels; then the most optimal needle lengths with the least total event risk were determined for each gender. Data were analyzed using SPSS software version 17 (SPSS Inc,Chicago, Illinois, USA). All measures were compared using the Mann-Whitney U-test. For all analyses, p<0.05 was considered statistically significant.

Study Population
The general characteristics of 208 patients are shown in Table I. When compared to men, women had lower height (159±8 vs. 174±8 cm; p<0.001) and higher BMI values (29.9±9 vs. 26.4±5.1 kg/cm 2 ; p<0.001). Table II shows the ultrasonographic measures and change rates during minimal and maximal pressures. STMD min , STMD max , and STBD max were significantly higher in women. When maximum pressure was performed by the ultrasound probe, the mean change in STMD was found to be 1.3±1.6 mm (10.4%), and in STBD was found to be 11.7±4.3 mm (26.9%). The mean change in STMD was significantly higher in women (1.8±1.9 vs. 0.7±0.8 mm; p<0.001).

Ultrasonographic Measures and Their Correlation with Other Anthropometric Measures
When we classified patients into 3 groups according to their BMI values (Group 1: 18.5-24.9 kg/m 2 ; group 2: 25-29.9 kg/m 2 and group 3: ≥30 kg/m 2 ), we found that when compared to men with the similar BMI cut-off values, women had longer STMD max and STMD min values.
According to univariate analysis, only gender and BMI were found to be the predictors of all four ultrasonic measures. Multivariate analysis revealed that BMI was an independent risk factor for STMD max , STMD min , STBD max and STBD min ; and gender was an independent risk factor for STMD max and STMD min .
When all parameters were taken into consideration, there was a strong correlation between STMD max and STMD min (r=0.971; p<0.001), STBD max and STBD min (r=0.874; p<0.001), and STBD max and STMD max (r=0.67; p<0.001) (Table III). Even though height had the highest correlation with STMD min and STMD max , this correlation was not present when each gender was analyzed separately. Finally, within all three anthropometric measures, BMI had significant correlation with STMD min , STMD max , STBD min , and STBD max for each gender (Table III).

Risks of Subcutaneous or Peri/Intraosseous Injection
When an AAI with 15.2 mm needle length was taken as the reference in women, it was found that STMD max of 45 women (38%) and STMD min of 58 women (49.2%) were longer than 15.2 mm, resulting in inadequate penetration into the intramuscular area ( Figure 1). Twenty-nine women (64.4%) with STMD max >15.2 mm had BMI≥30. Subcutaneous injection and peri/intraosseous injection risks with different needle lengths for each gender are given in Table IV. Only one man (1.1%) had a STMD max longer than 15.2 mm ( Figure 2). There was no risk of peri/ intraosseous injection with a 15.2 mm AAI needle length in women, whereas two men (0.9%) had peri/intraosseous injection risk with a 15.2 mm needle length.   number was similar with different needle lengths ranging from 12 to 18 mm (3.3% total event risk). Since the total event risk was very low within this needle length range, men were not further sub-grouped according to their anthropometric measurements.

Optimum Adrenaline Auto-Injector Needle Length for Women
In women, there was a 46% subcutaneous injection risk with 14 mm needle length, whereas, this risk was expected to decrease to 0.8% with 28 mm needle length (Table IV). However, peri/intraosseous injection risk was also expected to increase with longer needles.

Optimum Adrenaline Auto-Injector Needle Length for Men
For men, subcutaneous injection risk was very low with different needle lengths ranging from 12 to 28 mm. As expected, peri/intraosseous injection risk was expected to be higher with longer needles (Table IV). With needle lengths longer than 18 mm, the risk increment was significant (n=5 for 18 mm needle length vs. n=10 for 20 mm needle length; p=0.025). With needle lengths shorter than 12 mm, subcutaneous injection risk was expected to be significantly higher (n=5 for 12 mm needle length vs. n=9 for 10 mm needle length; p=0.046). Total event The lowest subcutaneous injection and peri/intraosseous injection risks were expected between the needle lengths 20 to 26 mm in women. Total event risks were significantly higher when the needle was longer than 26 mm or shorter than 20 mm [(n=20 for 26 mm needle length vs. n=32 for 28 mm needle length; p=0.002) and (n=28 for 18 mm needle length vs. n=17 for 20 mm needle length; p=0.001), respectively]. A minimum of 15 and a maximum of 20 total events were expected within this range.
Since the total event risk was still very high for women, the effect of anthropometric measures on the results was evaluated. Mean BMI was significantly higher in women with STMD max ≥15.2 mm compared to those with STMD max <15.2 (32.4±5.5 vs. 28.3±5.8 kg/m 2 ; p<0.001). BMI also had the highest correlation with STMD max . Therefore, women were divided into subgroups according to their BMIs. Subcutaneous injection and peri/intraosseous injection risks with different needle lengths and different BMI threshold values are given in Table V. When each of five different BMI threshold values was evaluated separately, for women with higher BMI, the optimal needle length with the lowest total event risk was 26   were obese (8). Tsai et al. found that 28% of women (19/69) and none of the men had STMD max >15.2 mm (10). They also showed that women with STMD max >15.2 mm had higher BMI compared with STMD max ≤15.2 mm (33±8 vs. 25±4 kg/m 2 , p<0.001). Another recent study using ultrasonographic measures also found that 33% and 10% of adults had subcutaneous injection risks with high-pressure AAI (with 15 mm needle length) and lowpressure AAI (with 23 mm needle length), respectively (15). In another study, the same investigators also showed that obese women had higher subcutaneous injection risk with Epipen® (17).
For the proper activation of trigger mechanisms of the AAIs, there is a certain amount of pressure that needs to be performed on the skin. This pressure compresses the subcutaneous tissue and makes it thinner. Therefore, while calculating the optimum needle length, this factor should be taken into consideration. As far as we know, ultrasound is the best imaging tool that can imitate and show the changes in the subcutaneous tissue at the same time by probe compression (19). The change in STMD under minimum and maximum pressures has been reported as 11% in literature (19), whereas in our study it was 10.4%. Another study measured the difference in subcutaneous tissue diameter between the minimum and maximum pressures and found 14.1 mm change (range 3.7 -31.0 mm) (15). Dreborg et al. presented the data that pressure applied to release the needle for different AAIs may vary and the mean pressures needed to trigger different devices varied between 15-28 newtons. Even though the ultrasound probe surface area is much larger than AAIs, we believe that the maximum pressures applied in our study would be enough, and probably much more higher than, to activate the trigger mechanisms and can be adapted in all commercially available AAIs (15).
Other than needle length and tissue compression, another factor affecting the depth of injection into the right compartment is the propulsion force of the AAI. Tightly coiled spring inside the Epipen® delivers a force of 22.7 pounds (20) which causes a mean maximum injection depth of 29.68±2.08 mm in ballistic gelatin (21). This propulsion force may enable the drug to reach wider areas than the needle length. However, Diacono et al. clearly presented that the deep fascia of the thigh prevents fluid traveling from the subcutaneous tissue into the underlying muscle in animal tissues (22). The results of this study showed that the whole needle orifice mm. For all five BMI threshold levels, total event number with a 26 mm needle length was lower than that with 24 mm and 28 mm, but the difference was not significant. Similarly, for women with lower BMI, the optimal needle length with the lowest total event risk was 20 mm. For all 5 BMI threshold levels, total event number with 20 mm needle length was lower than that with 18 mm and 22 mm, but the difference was not significant.
Our results showed that in women with BMI>28 kg/m 2 , with a 26 mm AAI needle length, subcutaneous injection risk was 0.8% and peri/intraosseous injection risk was 2.5%; whereas in women with BMI≤28 kg/m 2 , with a 20 mm AAI needle length, subcutaneous injection risk was 1.7% and peri/intraosseous injection risk was 1.7% (6.7% total events in all women). In women with BMI>32 kg/m 2 , with a 26 mm AAI needle length, subcutaneous injection risk was 0.8% and peri/intraosseous injection risk was 0%; whereas in women with BMI≤32 kg/m 2 , with a 20 mm AAI needle length, subcutaneous injection risk was 4.2%, peri/intraosseous injection risk was 1.7% (6.7% total events in all women). Considering 28 kg/m 2 as threshold level would increase peri/intraosseous injection risk but decrease subcutaneous injection risk; whereas taking 32 kg/ m 2 would do the opposite.

DISCUSSION
Our results highlight that the risk for inadequate IM delivery is higher in women, especially those with BMI>32 kg/m 2 . Total event risk was much lower in men compared to women. In addition, two different AAI needle lengths and BMI cut-off values had been identified for the lowest total event risk in women.
Since designing a study to evaluate the IM delivery ratio and bioavailability of adrenaline in patients with anaphylaxis in real-life settings would be unethical and probably not possible, the present study, like previous studies, was performed on healthy subjects. Previous studies aiming to present IM delivery rates of different commercially available AAIs have shown that most of the auto-injectors available on the market might fail to reach the right compartment due to insufficient needle lengths (8,10,12,(14)(15)(16)(17)(18). In a preliminary study where STMD was calculated on computed tomography scans, Song et al. found that mean STMD of women was nearly twice that of men (0.66±0.47 cm for men and 1.48±0.72 for women). STMD was longer than Epipen® autoinjector needle length in only one obese man, and 21 women (21%), 11 of whom of the AAI needles must pass completely into the muscle. Therefore, in previous similar studies, while calculating the possible IM delivery ratios over STMD measurements, the needle length was reduced by 2 mm when estimating the STMD max (14,15,17). Since subcutaneous and peri/ intraosseous injection risks are separately calculated for needle lengths from 14 to 28 mm, this reduction was not done in our study.

What Should be the Optimal Needle Length for Men and Women?
According to our results, with a 15.2 mm needle length, subcutaneous injection risk was 1.1%, peri/intraosseous injection risk was 2.2% (total event risk 3.3%) in men. Since this result was found to be the same for both 14 and 16 mm needles, when the 2 mm orifice length was also considered for fascial penetration, we may propose 16 mm needle length is optimal for men.
On the other hand, there is not a single optimal needle length for women. Compared to men with the same BMI, STMD was higher in women. With 15.2 mm long needle length, total event risk was 38% in women, and it was all formed by subcutaneous injection risk. Between 20-26 mm needle lengths, total event risk was expected to lower to 12.7%. Since the percentage was still high, we tried to find different optimal needle lengths for women with different BMIs, which had the highest correlation with all ultrasonographic measures. As shown in Table V, it was possible to lower the total event risk to 6.7% with a needle length of 26 mm for women with higher BMI (BMI>28 or 32 kg/m 2 ) and 20 mm for women with lower BMI (BMI≤28 or 32 kg/m 2 ). Even though taking these two BMI thresholds did not affect the total event possibility, since peri/intraosseous injection risk was lower with 32 kg/m 2 , we suggest the optimal needle length should be 20 mm for women with BMI≤32 kg/m 2 and 26 mm for women with BMI>32 kg/m 2 . However, there were not many patients with very extreme BMI levels in our study, therefore, we think needle length should be individualized for these patients.
Besides optimizing needle lengths, there are other subjects that should be improved in AAIs available on the market. Dreborg et al. showed that needle lengths may show variations between two devices of the same manufacturer, and this variation was greatest for Epipen® at 5 mm (17). Also, as we noted before, the needed trigger pressure may also vary among the same brand devices (19) and these variabilities should also be minimized by the manufacturers. Lastly, in a recent study, Duvauchelle et al. claimed that needle length or IM injection may not be absolutely required for auto-injector efficacy (23). They showed that in 11 of 12 overweight women in whom the administered AAI's needle length failed to reach the muscle, the magnitude of the first plasma peak of adrenaline in terms of Tmax and Cmax did not differ from that observed in normal-weight men. They also observed a sequential, immediate and larger second peak in these patients which resulted in a significantly higher total bioavailability than in normal-weight men. However, this study only included a limited number of healthy subjects. Decreased vascular tone, capillary leakage, and intravascular volume loss occur during anaphylaxis. These changes in micro and macrovascular environment may result in changes in the tissue vascularization. Therefore, it is controversial whether the findings of this study can be adapted to patients with anaphylaxis. Also, if this study is ever supported by similar studies in the future, an additional question emerges as to whether adrenaline should be administered subcutaneously.
The present study has some limitations. Even though measurements showed no significant difference in 30 randomly selected patients, interobserver variation may be present among the two radiologists. Secondly, maximum pressure was defined as the maximum pressure performed by the radiologist and tolerated by the patient; but the exact pressure amount was not measured. Thirdly, in real life, patients usually use the AAIs with their clothes on. Since standardization is not possible, the effect of clothes on subcutaneous tissue compression and the possible increase in the distance could not be evaluated. Fourthly, the surface area of the ultrasound probe is wider than that of AAIs, and it is not possible to estimate how this affects our results. Lastly, even though our results are similar to studies from different countries, we cannot generalize our findings for other populations for certain.
In conclusion, intramuscular injection of adrenaline with an effective dosage is very important for the efficient treatment in anaphylaxis. Since BMI may differ among people, individualization of AAI needle lengths according to certain BMI values may be appropriate. Our results show the optimal needle lengths with the lowest total event rate as 16 mm for men independent from BMI, as 20 mm for women with BMI≤32 kg/m 2 , and 26 mm for women with BMI>32 kg/m 2 -underlining the necessity of individualized AAIs according to these reference intervals.