One disadvantage of some drugs administered IV, such as diazepam, is pain during the injection, which at times is very distressing to patients. Pain on injection of diazepam is primarily attributed to the presence of propylene glycol as a vehicle in some of its formulations or the precipitation of diazepam at its site of infusion.
Many different methods have been proposed to reduce the incidence or intensity of this adverse effect, such as injection into a large vein, administrating local anesthetics, opioids (1), metoclopramide (2,3) or diphenhydramine (4), and diluting the injectant with a solvent (1). Metoclopramide is a weak local anesthetic in its own right (1). There are no published studies investigating the analgesic efficacy of metoclopramide in abolishing the pain associated with diazepam injection. Therefore, we conducted a randomized, prospective, double-blinded, placebo-controlled clinical trial to identify the incidence of pain on injection of diazepam and to compare the analgesic efficacy of IV metoclopramide with lidocaine pretreatment before diazepam injection.
After approval by local ethics committee, we obtained written informed consent from all subjects. We studied 159 patients (ASA status I and II), aged 20–70 yr old, undergoing various elective surgical procedures. Patients with a history of Parkinsonism or any type of drug abuse or those who had small veins, thrombophlebitis, or any painful lesion were excluded from the study. Patients with allergies to lidocaine or metoclopramide or patients who had already received premedication within the previous 12 h were also excluded. In the anesthetic room, a 20-gauge catheter was placed into the vein of the dorsum of the hand and attached to a normal saline infusion. No other drugs (e.g., antibiotics) were administered through the IV before the study medications. The patients were randomized by sequential allocation into one of three groups. Patients in Group 1 (n = 53) received 2 mL of normal saline (placebo) IV immediately before 0.1 mg/kg of diazepam premedication, and Group 2 (n = 53) received 20 mg of preservative-free lidocaine 1% (2 mL) immediately before 0.1 mg/kg of diazepam premedication. In Group 3, metoclopramide 10 mg (2 mL) was injected IV immediately before the identical dose of diazepam premedication. The rate of diazepam injection was 1 mg/s for all of the three groups. The diazepam used in this study was formulated with propylene glycol.
Tested drugs were administered via the same line of fluid infusion in the upper limb, and the IV fluid was stopped while the injection of the tested drug was being performed. All the studied drugs were stored at room temperature (23°C), and all of the injections were performed at the port immediately proximal to the IV catheter at a rate of 0.5 mL/s. The anesthesiologist evaluating the patient’s response was different from the one administrating the drug and was blinded to the study drugs. All of the drugs were injected via identical 2-mL syringes. The patient’s response was graded as follows: 0 = absolutely without pain, 1 = no spontaneous expression of pain, but on questioning, patient expresses the mild sensation of pain, 2 = mild spontaneous expression of pain whether by verbal expression, grimace, or by movement at wrist only, and 3 = remarkable expression of pain whether by crying or movement/withdrawal of the involved arm (elbow/shoulder). Any score other than 0 represented pain on injection.
Statistical analysis was performed by SPSS package (SPSS Inc, Chicago, IL). Demographic data were analyzed by χ2 and analysis of variance tests. The intensity of pain was analyzed by Kruskal-Wallis and Mann-Whitney tests, and the incidence of pain was analyzed by χ2 tests. P values <0.05 were considered statistically significant.
One patient from Group 1 and four patients from Group 2 were subsequently withdrawn from the study because their opioid addiction was uncovered later.
A total of 154 patients comprising 82 men and 72 women were studied. There was no significant distribution difference in age (P > 0.05), weight (P = 0.980), and ASA status (P = 0.570) among the three groups (Table 1). There was a significant distribution difference in sex (P = 0.015) among the three groups, but, as will be explained later, this difference did not generate any significant problem in our results.
This study demonstrated that the incidence of pain on IV injection (any score other than 0 represented pain on injection) of diazepam was 83% in the placebo group (Table 2). Group 2 showed a significantly less frequent incidence of pain than the saline (P < 0.000) and the metoclopramide (P = 0.002) groups as diazepam was injected (Table 2). Although there was no significant difference in the incidence of pain between Groups 1 and 3 (P = 0.093;Table 2), Group 3 showed significantly fewer patients with severe pain scores (score 2 and 3) than the saline group as diazepam was injected (P < 0.000).
The intensity of pain was different among the three groups (P = 0.000). Both treatment groups (2 and 3) experienced significantly reduced intensity of pain in comparison with Group 1 (placebo) (P = 0.000). The intensity of pain in Group 2 was also significantly less than in Group 3 (Table 2;P = 0.012). These findings were also true within each group of either men or women patients separately. In other words, the sex of the patients did not affect their responses to the intensity of pain. After completion of the study, no extrapyramidal disturbances were observed.
We found metoclopramide pretreatment to be effective in abolishing pain on the injection of diazepam. Our study showed that the IV lidocaine pretreatment had a greater efficacy in reducing the intensity of pain. In our trial, there was no significant relationship between the sex of the patients and their response to pain-reducing strategies.
Ganta and Fee (5) reported that pretreatment with lidocaine or metoclopramide immediately before propofol injection significantly reduced the incidence of pain compared with placebo. Their patients received variable oral benzodiazepine premedications, which may have influenced reports of discomfort (1). In the present study, a larger dose of metoclopramide (10 mg versus 5 mg) was administered before the diazepam injection, and the study was also performed in unpremedicated patients. Ganta and Fee (5) evaluated only the incidence of pain, but we demonstrated both the incidence and intensity of pain. In their study, the reduced incidence of pain was comparable within their two drug groups, and it was not in agreement with our observations regarding the incidence of pain. We were also eager to know whether the analgesic efficacy induced by metoclopramide was as effective as lidocaine. Although the metoclopramide was effective in abolishing pain on injection, the analgesic efficacy of our two study drugs was different. Lidocaine pretreatment was more effective.
Mecklem (6) compared the metoclopramide-propofol mixture with a lidocaine-propofol combination and observed that the incidence of injection pain was similar in both groups. Mixing lidocaine or metoclopramide with an injectant (propofol) may alleviate this problem, but there are no data regarding the chemical compatibility of the drugs. Incompatibility has been reported between diazepam and several other drugs, and its admixture with other drugs should be avoided.
Liaw et al. (3) observed that metoclopramide retention with the tourniquet is as effective as lidocaine and may be an alternative drug for analgesia with propofol injection. They also observed that metoclopramide mixed with propofol is less efficient than metoclopramide injected IV before the propofol injection (3).
It is yet to be determined whether the rubber retention technique would be safe and effective in reducing diazepam injection pain, and it remains to be established whether metoclopramide-induced analgesia on diazepam injection is dose-dependent.
Metoclopramide (2-methoxy-5-chloroprocainamide), primarily an antiemetic and a prokinetic drug, shares structural and physicochemical properties with lidocaine (6), procaine (3), and procainamide (3) and is a weak local anesthetic in its own right (1,6). Although metoclopramide, like morphine, may alter the influx of calcium ions across the membrane to produce a generalized analgesic effect, the exact mechanism whereby it prevents local pain is unknown (5). However, an investigation in patients undergoing second trimester abortion showed that morphine requirements were reduced significantly by IV metoclopramide (5). The drug reduces spasm in the fallopian tubes, and it may be on this basis that venous pain is attenuated (5,7). Metoclopramide also provides analgesia for ureteric colic (8). This analgesic effect of metoclopramide in the genitourinary tract may be attributed to its antagonism of the dopamine receptors, as well as its cholinergic activity, which reduces smooth muscle spasm and increases effective peristaltic action (7).
The limitations of this study may concern its focus on pain and thrombophlebitis after an IV diazepam administration. These may not occur for several days and thus may have been under reported. Metoclopramide administration is not recommended in patients with extrapyramidal disturbances.
Metoclopramide, rather than lidocaine, pretreatment may be a reasonable analgesic alternative for painful injections, especially when there is a medical condition where lidocaine should be used very cautiously.
The authors thank Dr. Ali Bijani for his assistance with statistical analysis. They are grateful to Hasan Arbab for his excellent suggestions and for typing the manuscript. The authors are also indebted to Jeanette Esau, Editorial Assistant of Anesthesia & Analgesia, for her patience and valuable cooperation.
1. Picard P, Tramer MR. Prevention of pain on injection with propofol: a quantitative systematic review. Anesth Analg 2000; 90: 963–9.
2. Maroof M, Khan RM, Khalid A, et al. Pain associated with propofol injection is abolished by pretreatment with metoclopramide. Br J Anaesth 1995; 74 (suppl 1): 8A.
3. Liaw WJ, Pang WW, Chang DP, Hwang MH. Pain on injection of propofol: the mitigating influence of metoclopramide using different techniques. Acta Anaesthesiol Scand 1999; 43: 24–7.
4. Green SM, Rothrock SG, Gorchynski J. Validation of diphenhydramine as a dermal local anesthetic. Ann Emerg Med 1994; 23: 1284–9.
5. Ganta R, Fee JPH. Pain on Injection of propofol: comparison of lignocaine with metoclopramide. Br J Anaesth 1992; 69: 316–7.
6. Mecklem DW. Propofol injection pain: comparing the addition of lignocaine or metoclopramide. Anaesth Intensive Care 1994; 22: 568–70.
7. Rosenblatt WH, Cioffi AM, Sinatra R, et al. Metoclopramide: an analgesic adjunct to patient-controlled analgesia. Anesth Analg 1991; 73: 553–5.
8. Muller TF, Naesh O, Svare E, et al. Metoclopramide (Primperan) in the treatment of ureterolithiasis: a prospective double-blind study of metoclopramide compared with morphatropin on ureteral colic. Urol Int 1990; 45: 112–3.