- Email: [email protected]
Comparison of effects of lidocaine hydrochloride, buffered lidocaine, diphenhydramine, and normal saline after intradermal injection
Original Contributions Comparison of Effects of Lidocaine Hydrochloride, Buffered Lidocaine, Diphenhydramine, and Normal Saline After Intradermal Injection Yun Xia, MD, PhD,* Edward Chen, MD,† David L. Tibbits, MD,† Thomas E. Reilley, DO,‡ Thomas D. McSweeney, BS§ Department of Anesthesiology, Division of Critical Care Medicine, The Ohio State University Medical Center, Columbus OH
*Assistant Professor of Anesthesiology †Clinical Instructor ‡Associate Professor of Anesthesiology and Associate Professor of Pharmacy §Data and Statistics Manager Address reprint requests and correspondence to Dr. Reilley at the Department of Anesthesiology, Division of Critical Care Medicine, The Ohio State University Medical Center, Doan Hall Room N416, 410 West 10th Ave., Columbus, OH 43210-1228, USA. E-mail: [email protected] Received for publication May 7, 2001; revised manuscript accepted for publication March 20, 2002.
Study Objective: To evaluate pain and the spread of analgesia when local anesthetics are given as an intradermal injection into the dorsal aspect of the hand. Design: Randomized, double-blinded, placebo-controlled study. Setting: University medical center. Patients: 40 consenting adult volunteers. Interventions: Volunteers were randomly assigned to receive a 0.25-mL injection of either lidocaine hydrochloride (1%), buffered lidocaine, diphenhydramine (1%), or placebo (0.9% sodium chloride solution) into the dorsal aspect of both hands. Measurements: The volunteers used a visual analog scale to compare the pain of needle insertion and solution injection. Then at 1, 2, 5, 10, 20, and 30 minutes after intradermal injection, the extent of the analgesic area was marked on a strip of tape placed horizontally across the hand. Then at 32 minutes after intradermal injection, the extent of the analgesic area was marked on a strip of tape placed vertically across the hand. The volunteers were called each day and asked the duration of their numbness or hyperesthesia until their hands were no longer numb or sore. Main Results : Buffered lidocaine during intradermal infiltration was found to be significantly (p ⬍ 0.05) less painful than either lidocaine hydrochloride or diphenhydramine and equivalent to placebo. Diphenhydramine and lidocaine hydrochloride during intradermal infiltration induced significantly (p ⬍ 0.05) more pain than buffered lidocaine or placebo. Lidocaine hydrochloride displayed a significantly (p ⬍ 0.05) larger diameter of analgesia than placebo by 1 minute after the injection, buffered lidocaine by 2 minutes after injection, and diphenhydramine by 5 minutes after injection. By 20 minutes after injection, diphenhydramine diameter of analgesia was significantly (p ⬍ 0.05) larger than placebo but significantly less than buffered lidocaine. By 30 minutes after injection, diphenhydramine diameter of analgesia was equivalent to placebo whereas buffered lidocaine and lidocaine diameters were still significantly (p ⬍ 0.05) larger than placebo. Diphenhydramine injection resulted in numbness that lasted significantly (p ⬍ 0.05)
Journal of Clinical Anesthesia 14:339 –343, 2002 © 2002 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
0952-8180/02/$–see front matter PII S0952-8180(02)00369-0
Original Contributions
longer than other study solutions whereas buffered lidocaine and lidocaine injections resulted in numbness that lasted significantly longer than placebo. Diphenhydramine injection resulted in hyperesthesia that lasted for 2 or more days in 12 of the volunteers. Conclusion: There is a reduction of infiltration pain using buffered lidocaine as opposed to lidocaine and diphenhydramine. Although lidocaine injection resulted in a slightly faster spread of analgesic diameter, buffered lidocaine was equivalent to lidocaine from minute 2 until minute 30. Therefore, to obtain optimal anesthetic conditions, we recommend that buffered lidocaine be given 2 minutes before performing catheterization, whereas diphenhydramine should be given 5 minutes before catheterization, but only when buffered lidocaine cannot be used. © 2002 by Elsevier Science Inc. Keywords: Anesthetics, local: diphenhydramine, lidocaine; drug injection, intradermal; lidocaine, buffered.
Introduction In almost all procedures performed in the operating room, at least one intravenous (IV) access should be established for a patient. For procedures such as coronary artery bypass graft surgery, several invasive monitors are used in addition to the peripheral IV access used during the operation, e.g., an arterial catheter and a pulmonary arterial catheter. It has been shown that intradermal injection of local anesthetics can reduce discomfort during insertion of the IV catheter and placement of the invasive monitors. 1–3 Lidocaine is the most commonly used local anesthetic in intradermal infiltration, topical anesthesia, and peripheral nerve blocks. Lidocaine is an amino amide with a pKa of 7.9, and it is generally provided in a solution with pH of 6.5.4 It is known that the addition of bicarbonate to local anesthetics increases the pH of the solution and thereby the percentage of the uncharged form of the local anesthetics, and is believed to provide faster onset and spread of the block.5 It is also widely believed among physicians that adding sodium bicarbonate to the lidocaine hydrochloride solution can minimize the painful sensation during intradermal injection of lidocaine.6 –10 However, double-blind human studies have not substantiated the hypothesis regarding a reduction of latency with lidocaine carbonate solutions during epidural anesthesia.11 Diphenhydramine has long been known for its antihistamine activity and its local analgesic property.12 When lidocaine is contraindicated because of hypersensitivity to lidocaine or to the preservatives in its solution, diphenhydramine has been used as a local anesthetic for intradermal infiltration.13 However, to date, no single study has compared diphenhydramine to lidocaine or buffered lidocaine for pain and spread of analgesia. We used sodium chloride (0.9%) solution as a placebo for our control group in our double-blind, randomized, study, which we designed to compare the differences in spread of analgesia and perceived pain after intradermal injection of local anesthetic solutions, including lidocaine 340
J. Clin. Anesth., vol. 14, August 2002
hydrochloride, buffered lidocaine, diphenhydramine, and 0.9% sodium chloride.
Materials and Methods After obtaining Ohio State University Medical Center Institutional Human Subjects Review Committee approval and written informed consent, 40 healthy adult volunteers were entered into our study, which was performed in a single day. Volunteers refrained from using antihistamine medications for 24 hours before participation. None of the volunteers was pregnant or lactating, had obstructive bladder disease, or had a known allergy or other adverse reaction to diphenhydramine or local anesthetics. Each of the 80 study syringes was filled with one of the following four solutions: 1% lidocaine hydrochloride with preservative (Elkins-Sinn, Inc., Cherry Hill, NJ), buffered lidocaine, 1% diphenhydramine (Warner-Lambert Co., Morris Plains, NJ), or 0.9% sodium chloride without the preservative benzyl alcohol (Baxter, Deerfield, IL). Buffered lidocaine was prepared by adding 8.4% sodium bicarbonate (Abbott Laboratories, North Chicago, IL) to lidocaine hydrochloride solution at a 1:9 ratio. The syringes were then numbered randomly from 1 to 80 and stored at room temperature. To ensure uniformity throughout the study, a single skilled resident (E.C.), who was blinded to content of each syringe, performed the intradermal injections. After cleaning the skin with 70% isopropyl alcohol pads, he performed each injection with a 26-gauge, 0.25-inch needle on the dorsal aspect of each hand of the 40 healthy volunteers. After each volunteer had used a 100-mm visual analog scale (VAS) to indicate the pain of needle insertion and reported that all pain had ceased from needle insertion, each hand (randomized arm) of the blinded volunteer received an injection of 0.25 mL of one of the study solutions. Injections were made over 2 seconds, and the volunteer again used a 100-mm VAS to indicate the pain of solution infiltration. After the infusion was completed and the needle withdrawn, a strip of tape was placed horizontally across the back of the hand perpendicular to the point of insertion. The subject was asked to mark on the tape the extent of analgesia in both directions from the point of injection at 1, 2, 5, 10, 20, and 30 minutes after infusion of the study solution. These marked points were then labeled with the numerals 1, 2, 5, 10, 20, or 30. Then, at 32 minutes after injection, the horizontal strip of tape was removed from the back of the hand and a second strip of tape was placed vertically on the back of the hand perpendicular to the first strip. The subject was then asked to mark on the tape the extent of analgesia in both directions from the point of injection. This tape was labeled “32” and then removed from the hand. The marked tapes were then placed on data collection sheets labeled with the syringe number. Just after the extent of analgesia was accessed in the first randomized hand 10 minutes after the injection, the second hand was injected with its randomized solution. The volunteers were telephoned each day and asked the duration of their numb-
Intradermal injection of local anesthetics: Xia et al.
ness or hyperesthesia until their hands were no longer numb or sore. Statistical analysis was performed using SigmaStat for Windows (V2.03, SPSS Inc., Chicago, IL). The nonparametric VAS data were analyzed using the Kruskal-Wallis analysis of variance (ANOVA) on ranks and subsequently by the Tukey test. The interval anesthetic diameter data and the duration of numbness data were first tested for normality by the Kolmogorov-Smirnov test. An estimate of the population’s standard deviation for the anesthetic diameter data were derived from the 24 standard deviations obtained during this study (four groups ⫻ six time periods). An estimate of the population’s standard deviation for the duration of numbness data were derived from the standard deviations of the buffered lidocaine and lidocaine numbness data. These estimated population standard deviations were used in power calculations. Anesthetic diameter data were first tested with two-way repeated measures ANOVA on the factors time and study drug with the Tukey test on all pairwise multiple comparisons. If there was a significant difference in study drug, time, and study drug/time interactions then multiple one-way ANOVAs would be performed to compare differences between study groups at each time point and for each study group over the time periods with the Tukey test on all pairwise multiple comparisons. All one-way ANOVAs were corrected for multiple comparisons by the Bonferroni multiple comparison method. Differences were not considered significant unless the differences were equal or larger than the minimal detectable difference determined by the power analysis, and if the Bonferroni corrected multiple comparison p-value was equal to or less than 0.05. Data in the text and figures are presented as counts or as means ⫾ standard deviations.
Results None of our volunteer subjects (22 male, 18 female) was aware of which local anesthetic study solution was given. All intradermal injections were successful on the first attempt. As expected, the Kruskal-Wallis one-way ANOVA on Ranks found that the VAS data on the pain of needle insertion into the dorsal aspect of the hand was equivalent for all study groups. Comparison of VAS data on study solution infusion pain indicated that there was a significant (p ⬍ 0.001) difference between study groups. The pain of infusing diphenhydramine or lidocaine solution was significantly (p ⬍ 0.001) greater than was the pain from infusing either buffered lidocaine or placebo (Figure 1). There was no difference in the perceived infusion pain for the diphenhydramine and lidocaine-treated volunteers. There also was no difference in the perceived infusion pain for the buffered lidocaine- and salinetreated volunteers. The Kolmogorov-Smirnov test found that the duration of numbness data were normally distributed. Power analysis of the duration of numbness data found that we could detect differences between study groups ⱖ33 minutes with an ␣-error of 5% and a -error of 20% (0.8 power). The duration of numbness was significantly (p ⬍ 0.01) shorter
Figure 1. Pain, as measured by subjective visual analog scale (VAS), of “insertion” of the 26-gauge, 0.25-inch needle into the dorsal aspect of the hand and the pain of “infiltration” of 0.25 mL of the study solution. a ⫽ significantly (p ⬍ 0.001) more pain than infusing buffered lidocaine or placebo.
in the placebo study group compared to the other study groups. The duration of numbness was significantly (p ⬍ 0.01) longer in the diphenhydramine study group than the other study groups (Figure 2). In addition, 60% (12/20) of the volunteers who received diphenhydramine had hyperesthesia that lasted for 2 or more days (3.1 ⫾ 2.1 day). The Kolmogorov-Smirnov test found that the analgesic diameter data were not normally distributed. It was found that a volunteer who was given diphenhydramine (syringe #72) skewed the normal distribution. When syringe #72 data were removed, the Kolmogorov-Smirnov test found that the remaining analgesic diameter data were normally distributed. Therefore, all statistical tests were performed on normally distributed data (without syringe #72). Power analysis of the analgesic diameter data found that we could detect differences between study groups ⱖ10.5 mm with an ␣-error of 5% and a -error of 20% (0.8 power). Two-way repeated measures ANOVA on the factors drug
Figure 2. The duration of numbness for each study group. a ⫽ a significantly longer duration of numbness (ⱖ 33 min) when compared with the placebo group volunteers. b ⫽ a significantly shorter duration of numbness when compared with the diphenhydramine group volunteers. J. Clin. Anesth., vol. 14, August 2002
341
Original Contributions
mained larger than the saline-treated volunteers up to 20 minutes after infusion of the study solution; however, by 30 minutes after infusion there was no difference in the analgesic area diameter between the diphenhydraminetreated volunteers and the saline-treated volunteers. As can be seen in Figure 3, the analgesic area diameter of the volunteers given buffered lidocaine was significantly (p ⬍ 0.02) larger than the analgesic area diameter of the volunteers given diphenhydramine. When we subtracted the horizontal spread of the analgesic area diameter (minute 30) from the vertical spread of the analgesic area diameter (minute 32) to measure the even spread of analgesia in all directions, we found little difference. The average difference between horizontal and vertical spread for the study groups were: lidocaine 6 ⫾ 5 mm, buffered lidocaine 7 ⫾ 9 mm, diphenhydramine –2 ⫾ 14 mm, and saline –1 ⫾ 4 mm.
Figure 3. The diameter of the analgesic area for each of the anesthetic solution study groups over time. a ⫽ values significantly different from saline volunteers with a difference in diameter of ⱖ10.5 mm. b ⫽ values significantly different from buffered lidocaine volunteers, with a difference in diameter of ⱖ10.5 mm. c ⫽ values significantly different from minute 1, with a difference in diameter of ⱖ10.5 mm.
and time found a significant (p ⬍ 0.001) difference between the drug groups, a significant (p ⬍ 0.001) change over time, and a significant (p ⬍ 0.02) interaction between drug and time. The Tukey multiple comparison test found that the analgesic area diameter for buffered lidocaine at minute 20 was significantly (p ⬍ 0.01) greater than analgesic area diameter at minute 1 and that this difference was greater than 10.5 mm. All other study groups’ changes in analgesic area diameter from minute 1 were less than 10.5 mm. Tukey multiple comparisons between study groups at each time point indicated that the lidocainetreated volunteers had an analgesic area diameter that was significantly (p ⬍ 0.002) larger than the analgesic area diameter of the saline-treated volunteers 1 minute after infusion of the study solution and this difference was greater than 10.5 mm. The analgesic area diameter of the lidocaine-treated volunteers remained larger than the saline-treated volunteers for the remainder of the study (Figure 3). Buffered lidocaine-treated volunteers had an analgesic area diameter that was significantly (p ⬍ 0.001) larger than the analgesic area diameter of the salinetreated volunteers 2 minutes after infusion of the study solution, and this difference was greater than 10.5 mm. The analgesic area diameter of the buffered lidocainetreated volunteers remained larger than the saline-treated volunteers for the remainder of the study. Volunteers treated with diphenhydramine had an analgesic area diameter that was significantly (p ⬍ 0.005) larger than the analgesic area diameter of the saline-treated volunteers 5 minutes after infusion of the study solution, and this difference was greater than 10.5 mm. The analgesic area diameter of the diphenhydramine-treated volunteers re342
J. Clin. Anesth., vol. 14, August 2002
Discussion Intradermal injection of local anesthetics is commonly used to provide anesthesia for catheter placement. It has been shown that prior injection of a local anesthetic solution significantly reduced the pain associated with IV catheter insertion, and the pain of local anesthetic injection was much less than that of inserting an IV catheter without previous local anesthetic analgesia.3 The VAS is a common tool for measuring pain from local anesthetics.2,3,12–14 This study is, to our knowledge, the first to compare diphenhydramine, normal saline, lidocaine, and buffered lidocaine in a single group of volunteers. Our study found that the pain of local anesthetic infiltration of lidocaine was similar to the pain of diphenhydramine infiltration. Our results differ from those of Nuttall et al. 3 and Green et al.,12 whose studies used twice our infiltration volume (0.5 mL). Our study found that the pain of local anesthetic infiltration of lidocaine was greater than the pain of buffered lidocaine. These results are confirmed by others3,6 –10 who found that buffered lidocaine caused significantly less pain during infiltration than did lidocaine containing preservative. In addition, we found that there was no significant difference on perceived pain between buffered lidocaine and normal saline infiltration. The Nuttall et al.3 study found similar results; however, McKay et al.6 found that buffered lidocaine caused less pain than did normal saline infiltration. We also found that the pain of diphenhydramine (1%) infiltration was significantly greater than the infiltration pain of the other tested local anesthetics. Other studies also have found that diphenhydramine infiltration is painful.12–14 Our results show that the diphenhydramine-treated volunteers had considerably longer duration of numbness compared with all other study groups. The duration of numbness in the lidocaine, buffered lidocaine, and the saline-treated volunteers was less than 1.5 hours. Although diphenhydramine (1%) did produce effective local anesthesia, the disadvantages of severe pain on injection, prolonged analgesia, and prolonged rebound hyperesthesia make this drug an impractical choice, particularly when benzyl alcohol, which is another alternative to lidocaine,
Intradermal injection of local anesthetics: Xia et al.
has been shown to be effective in producing analgesia without the prolonged numbness.2 The spread of the analgesic diameter over time was very similar for the lidocaine- and buffered lidocaine-treated volunteers. There is a reduction of infiltration pain using buffered lidocaine compared with lidocaine and diphenhydramine. Although lidocaine injection resulted in a slightly faster spread of analgesic diameter, buffered lidocaine was equivalent to lidocaine from minute 2 until minute 30. Therefore, to obtain optimal anesthetic conditions, we recommend that buffered lidocaine be given 2 minutes before performing catheterization, whereas diphenhydramine should be given 5 minutes before catheterization, but only when buffered lidocaine cannot be used.
Acknowledgments The authors wish to gratefully acknowledge our 40 volunteers without whom this study would not have been accomplished.
References 1. Tyler IL: The relative pain inflicted by techniques for insertion of needles [Letter]. Anesth Analg 1984;63:373– 4. 2. McNelis KA: Intradermal bacteriostatic 0.9% sodium chloride containing the preservative benzyl alcohol compared with intradermal lidocaine hydrochloride 1% for attenuation of intravenous cannulation pain. AANA J 1998;66:583–5. 3. Nuttall GA, Barnett MR, Smith RL 2nd, Blue TK, Clark KR, Payton BW: Establishing intravenous access: a study of local anesthetic efficacy. Anesth Analg 1993;77:950 –3.
4. Carpenter RL, Mackey DC: Local anesthetics. In: Barash PG, Cullen BF, Stoelting RK (eds): Clinical Anesthesia, 3d ed. Philadelphia: Lippincott Williams & Wilkins, 1996:413– 40. 5. Williams MJ: Local anesthetics. In: Raj PP (ed): Pain Medicine. Chicago: Mosby, 1996:162–75. 6. McKay W, Morris R, Mushlin P: Sodium bicarbonate attenuates pain on skin infiltration with lidocaine, with or without epinephrine. Anesth Analg 1987;66:572– 4. 7. Christoph RA, Buchanan L, Begalla K, Schwartz S: Pain reduction in local anesthetic administration through pH buffering. Ann Emerg Med 1988;17:117–20. 8. Bartfield JM, Gennis P, Barbera J, Breuer B, Gallagher EJ: Buffered versus plain lidocaine as a local anesthetic for simple laceration repair. Ann Emerg Med 1990;19:1387–9. 9. Matsumoto AH, Reifsnyder AC, Hartwell GD, Angle JF, Selby JB Jr, Tegtmeyer CJ: Reducing the discomfort of lidocaine administration through pH buffering. J Vasc Interv Radiol 1994;5:171–5. 10. Colaric KB, Overton DT, Moore K: Pain reduction in lidocaine administration through buffering and warming. Am J Emerg Med 1998;16:353– 6. 11. Morison DH: A double-blind comparison of carbonated lidocaine and lidocaine hydrochloride in epidural anaesthesia. Can Anaesth Soc J 1981;28:387–9. 12. Green SM, Rothrock SG, Gorchynski J: Validation of diphenhydramine as a dermal local anesthetic. Ann Emerg Med 1994;23: 1284 –9. 13. Uckan S, Guler N, Sumer M., Ungor M: Local anesthetic efficacy for oral surgery: comparison of diphenhydramine and prilocaine. Oral Surg Oral Med Pathol Oral Radiol Endod 1998;86:26 –30. 14. Bartfield JM, Jandreau SW, Raccio-Robak N: Randomized trial of diphenhydramine versus benzyl alcohol with epinephrine as an alternative to lidocaine local anesthesia. Ann Emerg Med 1998;32: 650 – 4.
J. Clin. Anesth., vol. 14, August 2002
343
*Assistant Professor of Anesthesiology †Clinical Instructor ‡Associate Professor of Anesthesiology and Associate Professor of Pharmacy §Data and Statistics Manager Address reprint requests and correspondence to Dr. Reilley at the Department of Anesthesiology, Division of Critical Care Medicine, The Ohio State University Medical Center, Doan Hall Room N416, 410 West 10th Ave., Columbus, OH 43210-1228, USA. E-mail: [email protected] Received for publication May 7, 2001; revised manuscript accepted for publication March 20, 2002.
Study Objective: To evaluate pain and the spread of analgesia when local anesthetics are given as an intradermal injection into the dorsal aspect of the hand. Design: Randomized, double-blinded, placebo-controlled study. Setting: University medical center. Patients: 40 consenting adult volunteers. Interventions: Volunteers were randomly assigned to receive a 0.25-mL injection of either lidocaine hydrochloride (1%), buffered lidocaine, diphenhydramine (1%), or placebo (0.9% sodium chloride solution) into the dorsal aspect of both hands. Measurements: The volunteers used a visual analog scale to compare the pain of needle insertion and solution injection. Then at 1, 2, 5, 10, 20, and 30 minutes after intradermal injection, the extent of the analgesic area was marked on a strip of tape placed horizontally across the hand. Then at 32 minutes after intradermal injection, the extent of the analgesic area was marked on a strip of tape placed vertically across the hand. The volunteers were called each day and asked the duration of their numbness or hyperesthesia until their hands were no longer numb or sore. Main Results : Buffered lidocaine during intradermal infiltration was found to be significantly (p ⬍ 0.05) less painful than either lidocaine hydrochloride or diphenhydramine and equivalent to placebo. Diphenhydramine and lidocaine hydrochloride during intradermal infiltration induced significantly (p ⬍ 0.05) more pain than buffered lidocaine or placebo. Lidocaine hydrochloride displayed a significantly (p ⬍ 0.05) larger diameter of analgesia than placebo by 1 minute after the injection, buffered lidocaine by 2 minutes after injection, and diphenhydramine by 5 minutes after injection. By 20 minutes after injection, diphenhydramine diameter of analgesia was significantly (p ⬍ 0.05) larger than placebo but significantly less than buffered lidocaine. By 30 minutes after injection, diphenhydramine diameter of analgesia was equivalent to placebo whereas buffered lidocaine and lidocaine diameters were still significantly (p ⬍ 0.05) larger than placebo. Diphenhydramine injection resulted in numbness that lasted significantly (p ⬍ 0.05)
Journal of Clinical Anesthesia 14:339 –343, 2002 © 2002 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
0952-8180/02/$–see front matter PII S0952-8180(02)00369-0
Original Contributions
longer than other study solutions whereas buffered lidocaine and lidocaine injections resulted in numbness that lasted significantly longer than placebo. Diphenhydramine injection resulted in hyperesthesia that lasted for 2 or more days in 12 of the volunteers. Conclusion: There is a reduction of infiltration pain using buffered lidocaine as opposed to lidocaine and diphenhydramine. Although lidocaine injection resulted in a slightly faster spread of analgesic diameter, buffered lidocaine was equivalent to lidocaine from minute 2 until minute 30. Therefore, to obtain optimal anesthetic conditions, we recommend that buffered lidocaine be given 2 minutes before performing catheterization, whereas diphenhydramine should be given 5 minutes before catheterization, but only when buffered lidocaine cannot be used. © 2002 by Elsevier Science Inc. Keywords: Anesthetics, local: diphenhydramine, lidocaine; drug injection, intradermal; lidocaine, buffered.
Introduction In almost all procedures performed in the operating room, at least one intravenous (IV) access should be established for a patient. For procedures such as coronary artery bypass graft surgery, several invasive monitors are used in addition to the peripheral IV access used during the operation, e.g., an arterial catheter and a pulmonary arterial catheter. It has been shown that intradermal injection of local anesthetics can reduce discomfort during insertion of the IV catheter and placement of the invasive monitors. 1–3 Lidocaine is the most commonly used local anesthetic in intradermal infiltration, topical anesthesia, and peripheral nerve blocks. Lidocaine is an amino amide with a pKa of 7.9, and it is generally provided in a solution with pH of 6.5.4 It is known that the addition of bicarbonate to local anesthetics increases the pH of the solution and thereby the percentage of the uncharged form of the local anesthetics, and is believed to provide faster onset and spread of the block.5 It is also widely believed among physicians that adding sodium bicarbonate to the lidocaine hydrochloride solution can minimize the painful sensation during intradermal injection of lidocaine.6 –10 However, double-blind human studies have not substantiated the hypothesis regarding a reduction of latency with lidocaine carbonate solutions during epidural anesthesia.11 Diphenhydramine has long been known for its antihistamine activity and its local analgesic property.12 When lidocaine is contraindicated because of hypersensitivity to lidocaine or to the preservatives in its solution, diphenhydramine has been used as a local anesthetic for intradermal infiltration.13 However, to date, no single study has compared diphenhydramine to lidocaine or buffered lidocaine for pain and spread of analgesia. We used sodium chloride (0.9%) solution as a placebo for our control group in our double-blind, randomized, study, which we designed to compare the differences in spread of analgesia and perceived pain after intradermal injection of local anesthetic solutions, including lidocaine 340
J. Clin. Anesth., vol. 14, August 2002
hydrochloride, buffered lidocaine, diphenhydramine, and 0.9% sodium chloride.
Materials and Methods After obtaining Ohio State University Medical Center Institutional Human Subjects Review Committee approval and written informed consent, 40 healthy adult volunteers were entered into our study, which was performed in a single day. Volunteers refrained from using antihistamine medications for 24 hours before participation. None of the volunteers was pregnant or lactating, had obstructive bladder disease, or had a known allergy or other adverse reaction to diphenhydramine or local anesthetics. Each of the 80 study syringes was filled with one of the following four solutions: 1% lidocaine hydrochloride with preservative (Elkins-Sinn, Inc., Cherry Hill, NJ), buffered lidocaine, 1% diphenhydramine (Warner-Lambert Co., Morris Plains, NJ), or 0.9% sodium chloride without the preservative benzyl alcohol (Baxter, Deerfield, IL). Buffered lidocaine was prepared by adding 8.4% sodium bicarbonate (Abbott Laboratories, North Chicago, IL) to lidocaine hydrochloride solution at a 1:9 ratio. The syringes were then numbered randomly from 1 to 80 and stored at room temperature. To ensure uniformity throughout the study, a single skilled resident (E.C.), who was blinded to content of each syringe, performed the intradermal injections. After cleaning the skin with 70% isopropyl alcohol pads, he performed each injection with a 26-gauge, 0.25-inch needle on the dorsal aspect of each hand of the 40 healthy volunteers. After each volunteer had used a 100-mm visual analog scale (VAS) to indicate the pain of needle insertion and reported that all pain had ceased from needle insertion, each hand (randomized arm) of the blinded volunteer received an injection of 0.25 mL of one of the study solutions. Injections were made over 2 seconds, and the volunteer again used a 100-mm VAS to indicate the pain of solution infiltration. After the infusion was completed and the needle withdrawn, a strip of tape was placed horizontally across the back of the hand perpendicular to the point of insertion. The subject was asked to mark on the tape the extent of analgesia in both directions from the point of injection at 1, 2, 5, 10, 20, and 30 minutes after infusion of the study solution. These marked points were then labeled with the numerals 1, 2, 5, 10, 20, or 30. Then, at 32 minutes after injection, the horizontal strip of tape was removed from the back of the hand and a second strip of tape was placed vertically on the back of the hand perpendicular to the first strip. The subject was then asked to mark on the tape the extent of analgesia in both directions from the point of injection. This tape was labeled “32” and then removed from the hand. The marked tapes were then placed on data collection sheets labeled with the syringe number. Just after the extent of analgesia was accessed in the first randomized hand 10 minutes after the injection, the second hand was injected with its randomized solution. The volunteers were telephoned each day and asked the duration of their numb-
Intradermal injection of local anesthetics: Xia et al.
ness or hyperesthesia until their hands were no longer numb or sore. Statistical analysis was performed using SigmaStat for Windows (V2.03, SPSS Inc., Chicago, IL). The nonparametric VAS data were analyzed using the Kruskal-Wallis analysis of variance (ANOVA) on ranks and subsequently by the Tukey test. The interval anesthetic diameter data and the duration of numbness data were first tested for normality by the Kolmogorov-Smirnov test. An estimate of the population’s standard deviation for the anesthetic diameter data were derived from the 24 standard deviations obtained during this study (four groups ⫻ six time periods). An estimate of the population’s standard deviation for the duration of numbness data were derived from the standard deviations of the buffered lidocaine and lidocaine numbness data. These estimated population standard deviations were used in power calculations. Anesthetic diameter data were first tested with two-way repeated measures ANOVA on the factors time and study drug with the Tukey test on all pairwise multiple comparisons. If there was a significant difference in study drug, time, and study drug/time interactions then multiple one-way ANOVAs would be performed to compare differences between study groups at each time point and for each study group over the time periods with the Tukey test on all pairwise multiple comparisons. All one-way ANOVAs were corrected for multiple comparisons by the Bonferroni multiple comparison method. Differences were not considered significant unless the differences were equal or larger than the minimal detectable difference determined by the power analysis, and if the Bonferroni corrected multiple comparison p-value was equal to or less than 0.05. Data in the text and figures are presented as counts or as means ⫾ standard deviations.
Results None of our volunteer subjects (22 male, 18 female) was aware of which local anesthetic study solution was given. All intradermal injections were successful on the first attempt. As expected, the Kruskal-Wallis one-way ANOVA on Ranks found that the VAS data on the pain of needle insertion into the dorsal aspect of the hand was equivalent for all study groups. Comparison of VAS data on study solution infusion pain indicated that there was a significant (p ⬍ 0.001) difference between study groups. The pain of infusing diphenhydramine or lidocaine solution was significantly (p ⬍ 0.001) greater than was the pain from infusing either buffered lidocaine or placebo (Figure 1). There was no difference in the perceived infusion pain for the diphenhydramine and lidocaine-treated volunteers. There also was no difference in the perceived infusion pain for the buffered lidocaine- and salinetreated volunteers. The Kolmogorov-Smirnov test found that the duration of numbness data were normally distributed. Power analysis of the duration of numbness data found that we could detect differences between study groups ⱖ33 minutes with an ␣-error of 5% and a -error of 20% (0.8 power). The duration of numbness was significantly (p ⬍ 0.01) shorter
Figure 1. Pain, as measured by subjective visual analog scale (VAS), of “insertion” of the 26-gauge, 0.25-inch needle into the dorsal aspect of the hand and the pain of “infiltration” of 0.25 mL of the study solution. a ⫽ significantly (p ⬍ 0.001) more pain than infusing buffered lidocaine or placebo.
in the placebo study group compared to the other study groups. The duration of numbness was significantly (p ⬍ 0.01) longer in the diphenhydramine study group than the other study groups (Figure 2). In addition, 60% (12/20) of the volunteers who received diphenhydramine had hyperesthesia that lasted for 2 or more days (3.1 ⫾ 2.1 day). The Kolmogorov-Smirnov test found that the analgesic diameter data were not normally distributed. It was found that a volunteer who was given diphenhydramine (syringe #72) skewed the normal distribution. When syringe #72 data were removed, the Kolmogorov-Smirnov test found that the remaining analgesic diameter data were normally distributed. Therefore, all statistical tests were performed on normally distributed data (without syringe #72). Power analysis of the analgesic diameter data found that we could detect differences between study groups ⱖ10.5 mm with an ␣-error of 5% and a -error of 20% (0.8 power). Two-way repeated measures ANOVA on the factors drug
Figure 2. The duration of numbness for each study group. a ⫽ a significantly longer duration of numbness (ⱖ 33 min) when compared with the placebo group volunteers. b ⫽ a significantly shorter duration of numbness when compared with the diphenhydramine group volunteers. J. Clin. Anesth., vol. 14, August 2002
341
Original Contributions
mained larger than the saline-treated volunteers up to 20 minutes after infusion of the study solution; however, by 30 minutes after infusion there was no difference in the analgesic area diameter between the diphenhydraminetreated volunteers and the saline-treated volunteers. As can be seen in Figure 3, the analgesic area diameter of the volunteers given buffered lidocaine was significantly (p ⬍ 0.02) larger than the analgesic area diameter of the volunteers given diphenhydramine. When we subtracted the horizontal spread of the analgesic area diameter (minute 30) from the vertical spread of the analgesic area diameter (minute 32) to measure the even spread of analgesia in all directions, we found little difference. The average difference between horizontal and vertical spread for the study groups were: lidocaine 6 ⫾ 5 mm, buffered lidocaine 7 ⫾ 9 mm, diphenhydramine –2 ⫾ 14 mm, and saline –1 ⫾ 4 mm.
Figure 3. The diameter of the analgesic area for each of the anesthetic solution study groups over time. a ⫽ values significantly different from saline volunteers with a difference in diameter of ⱖ10.5 mm. b ⫽ values significantly different from buffered lidocaine volunteers, with a difference in diameter of ⱖ10.5 mm. c ⫽ values significantly different from minute 1, with a difference in diameter of ⱖ10.5 mm.
and time found a significant (p ⬍ 0.001) difference between the drug groups, a significant (p ⬍ 0.001) change over time, and a significant (p ⬍ 0.02) interaction between drug and time. The Tukey multiple comparison test found that the analgesic area diameter for buffered lidocaine at minute 20 was significantly (p ⬍ 0.01) greater than analgesic area diameter at minute 1 and that this difference was greater than 10.5 mm. All other study groups’ changes in analgesic area diameter from minute 1 were less than 10.5 mm. Tukey multiple comparisons between study groups at each time point indicated that the lidocainetreated volunteers had an analgesic area diameter that was significantly (p ⬍ 0.002) larger than the analgesic area diameter of the saline-treated volunteers 1 minute after infusion of the study solution and this difference was greater than 10.5 mm. The analgesic area diameter of the lidocaine-treated volunteers remained larger than the saline-treated volunteers for the remainder of the study (Figure 3). Buffered lidocaine-treated volunteers had an analgesic area diameter that was significantly (p ⬍ 0.001) larger than the analgesic area diameter of the salinetreated volunteers 2 minutes after infusion of the study solution, and this difference was greater than 10.5 mm. The analgesic area diameter of the buffered lidocainetreated volunteers remained larger than the saline-treated volunteers for the remainder of the study. Volunteers treated with diphenhydramine had an analgesic area diameter that was significantly (p ⬍ 0.005) larger than the analgesic area diameter of the saline-treated volunteers 5 minutes after infusion of the study solution, and this difference was greater than 10.5 mm. The analgesic area diameter of the diphenhydramine-treated volunteers re342
J. Clin. Anesth., vol. 14, August 2002
Discussion Intradermal injection of local anesthetics is commonly used to provide anesthesia for catheter placement. It has been shown that prior injection of a local anesthetic solution significantly reduced the pain associated with IV catheter insertion, and the pain of local anesthetic injection was much less than that of inserting an IV catheter without previous local anesthetic analgesia.3 The VAS is a common tool for measuring pain from local anesthetics.2,3,12–14 This study is, to our knowledge, the first to compare diphenhydramine, normal saline, lidocaine, and buffered lidocaine in a single group of volunteers. Our study found that the pain of local anesthetic infiltration of lidocaine was similar to the pain of diphenhydramine infiltration. Our results differ from those of Nuttall et al. 3 and Green et al.,12 whose studies used twice our infiltration volume (0.5 mL). Our study found that the pain of local anesthetic infiltration of lidocaine was greater than the pain of buffered lidocaine. These results are confirmed by others3,6 –10 who found that buffered lidocaine caused significantly less pain during infiltration than did lidocaine containing preservative. In addition, we found that there was no significant difference on perceived pain between buffered lidocaine and normal saline infiltration. The Nuttall et al.3 study found similar results; however, McKay et al.6 found that buffered lidocaine caused less pain than did normal saline infiltration. We also found that the pain of diphenhydramine (1%) infiltration was significantly greater than the infiltration pain of the other tested local anesthetics. Other studies also have found that diphenhydramine infiltration is painful.12–14 Our results show that the diphenhydramine-treated volunteers had considerably longer duration of numbness compared with all other study groups. The duration of numbness in the lidocaine, buffered lidocaine, and the saline-treated volunteers was less than 1.5 hours. Although diphenhydramine (1%) did produce effective local anesthesia, the disadvantages of severe pain on injection, prolonged analgesia, and prolonged rebound hyperesthesia make this drug an impractical choice, particularly when benzyl alcohol, which is another alternative to lidocaine,
Intradermal injection of local anesthetics: Xia et al.
has been shown to be effective in producing analgesia without the prolonged numbness.2 The spread of the analgesic diameter over time was very similar for the lidocaine- and buffered lidocaine-treated volunteers. There is a reduction of infiltration pain using buffered lidocaine compared with lidocaine and diphenhydramine. Although lidocaine injection resulted in a slightly faster spread of analgesic diameter, buffered lidocaine was equivalent to lidocaine from minute 2 until minute 30. Therefore, to obtain optimal anesthetic conditions, we recommend that buffered lidocaine be given 2 minutes before performing catheterization, whereas diphenhydramine should be given 5 minutes before catheterization, but only when buffered lidocaine cannot be used.
Acknowledgments The authors wish to gratefully acknowledge our 40 volunteers without whom this study would not have been accomplished.
References 1. Tyler IL: The relative pain inflicted by techniques for insertion of needles [Letter]. Anesth Analg 1984;63:373– 4. 2. McNelis KA: Intradermal bacteriostatic 0.9% sodium chloride containing the preservative benzyl alcohol compared with intradermal lidocaine hydrochloride 1% for attenuation of intravenous cannulation pain. AANA J 1998;66:583–5. 3. Nuttall GA, Barnett MR, Smith RL 2nd, Blue TK, Clark KR, Payton BW: Establishing intravenous access: a study of local anesthetic efficacy. Anesth Analg 1993;77:950 –3.
4. Carpenter RL, Mackey DC: Local anesthetics. In: Barash PG, Cullen BF, Stoelting RK (eds): Clinical Anesthesia, 3d ed. Philadelphia: Lippincott Williams & Wilkins, 1996:413– 40. 5. Williams MJ: Local anesthetics. In: Raj PP (ed): Pain Medicine. Chicago: Mosby, 1996:162–75. 6. McKay W, Morris R, Mushlin P: Sodium bicarbonate attenuates pain on skin infiltration with lidocaine, with or without epinephrine. Anesth Analg 1987;66:572– 4. 7. Christoph RA, Buchanan L, Begalla K, Schwartz S: Pain reduction in local anesthetic administration through pH buffering. Ann Emerg Med 1988;17:117–20. 8. Bartfield JM, Gennis P, Barbera J, Breuer B, Gallagher EJ: Buffered versus plain lidocaine as a local anesthetic for simple laceration repair. Ann Emerg Med 1990;19:1387–9. 9. Matsumoto AH, Reifsnyder AC, Hartwell GD, Angle JF, Selby JB Jr, Tegtmeyer CJ: Reducing the discomfort of lidocaine administration through pH buffering. J Vasc Interv Radiol 1994;5:171–5. 10. Colaric KB, Overton DT, Moore K: Pain reduction in lidocaine administration through buffering and warming. Am J Emerg Med 1998;16:353– 6. 11. Morison DH: A double-blind comparison of carbonated lidocaine and lidocaine hydrochloride in epidural anaesthesia. Can Anaesth Soc J 1981;28:387–9. 12. Green SM, Rothrock SG, Gorchynski J: Validation of diphenhydramine as a dermal local anesthetic. Ann Emerg Med 1994;23: 1284 –9. 13. Uckan S, Guler N, Sumer M., Ungor M: Local anesthetic efficacy for oral surgery: comparison of diphenhydramine and prilocaine. Oral Surg Oral Med Pathol Oral Radiol Endod 1998;86:26 –30. 14. Bartfield JM, Jandreau SW, Raccio-Robak N: Randomized trial of diphenhydramine versus benzyl alcohol with epinephrine as an alternative to lidocaine local anesthesia. Ann Emerg Med 1998;32: 650 – 4.
J. Clin. Anesth., vol. 14, August 2002
343