An analysis of pain and analgesia after Mohs micrographic surgery

DERMATOLOGIC

SURGERY

An analysis of pain and analgesia after Mohs micrographic surgery Bahar F. Firoz, MD, MPH,a,b Leonard H. Goldberg, MD,a,b Ofer Arnon, MD,d and Adam J. Mamelak, MD, FRCPCc Houston, Texas; Mississauga, Ontario, Canada; and Beer-Sheva, Israel Background: Pain characteristics and analgesia in patients undergoing Mohs micrographic surgery have not been systematically studied. It is important to know about pain after Mohs micrographic surgery to better serve patient needs. Objective: We sought to measure pain in patients after Mohs micrographic surgery, and to investigate the relationship among postoperative pain, surgical characteristics, patient characteristics, and analgesics used. Methods: The Wong-Baker 0-to-10 pain scale was prospectively administered postoperatively to all patients presenting for Mohs micrographic surgery in a private practice setting between October 1, 2007, and December 31, 2008. Patients recorded their pain level from the day of surgery through postoperative day 4. The age, sex, location of surgery, number of lesions operated on, postoperative size, type of repair, severity of pain, and oral analgesics consumed and dosages used were recorded. Results: A total of 433 patients were included in the final analysis. The highest pain scores were found on the day of surgery and steadily declined until postoperative day 4 (P \ .000). In all, 52% of patients took pain medication on the day of surgery, which declined successively with each postoperative day. The highest mean pain scores were statistically significantly associated with repair type (flaps), age (\66 years), number of lesions, and consumption of narcotics for pain relief. No statistically significant differences existed for sex or postoperative defect size. Limitations: The instrument used to measure pain relied on patient self-report in a private practice surgery center. Only the validated Wong-Baker pain scale was used to assess pain in this study. Conclusion: Approximately half of the patients after Mohs micrographic surgery take medication for pain control. Type of closure, location of surgery, age, and type of pain medication taken were significantly associated with postoperative pain. ( J Am Acad Dermatol 2010;63:79-86.) Key words: analgesia; Mohs micrographic surgery; pain; postoperative pain; Wong-Baker pain scale.

T

o better serve patients’ needs, it is important to know approximately how many patients experience pain after surgery, and what factors affect postoperative pain. The movement to

From DermSurgery Associates, Houstona; Department of Dermatology, Weill Cornell Medical College, Methodist Hospital, Houstonb; Dermatology Consultants, Mississaugac; and Department of Plastic and Reconstructive Surgery, Soroka University Medical Center, Beer-Sheva.d Funding sources: None. Conflicts of interest: None declared. Reprint requests: Leonard H. Goldberg, MD, DermSurgery Associates, 7515 S Main St, Suite 240, Houston TX 77030. E-mail: [email protected]. 0190-9622/$36.00 ª 2009 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2009.10.049

assess and treat patient pain has become an integral part of every in-patient hospital stay. Pain control is part of the standard of care for all patients admitted to the hospital. This movement has largely been supported by The Joint Commission, a nonprofit organization that accredits and certifies more than 15,000 health care organizations and programs in the United States.1 This organization has required that a pain score be part of the vital signs assessment. This movement has started to involve outpatient medical and surgical subspecialties in addition to inpatient hospital admissions. Several subspecialties have started to systematically investigate postsurgical pain and its response to analgesics.2 One of the most commonly used pain scales is the Wong-Baker pain scale originally developed in the pediatric population.3,4 It is a series of faces 79

80 Firoz et al

J AM ACAD DERMATOL JULY 2010

expressing different degrees of pain and was subseabdomen. Closure types were complex and interquently validated in adults. The only study to date in mediate linear, granulate and epithelialize (G&E), which pain scales were administered to patients full-thickness skin graft, split-thickness skin graft, during Mohs micrographic surgery involves pain advancement, rotation, transposition, and island surrounding administration of different concentrapedicle flap. Pain medications listed included tions of local anesthesia.5 We initiated a prospective 500 mg of acetaminophen, 100 mg propoxyphene pain assessment study to all patients having Mohs napsylate and 650 mg acetaminophen or 5 mg micrographic surgery and inhydrocodone and 500 mg vestigated tumor location, acetaminophen; some paCAPSULE SUMMARY type of repair, size of defect, tients also took medications age, sex, and pain medicawe did not prescribe includThis was a prospective study that tions consumed after Mohs ing aspirin, tramadol, and investigated postoperative pain and micrographic surgery. NSAIDs. These medications analgesic use after Mohs micrographic were included in the surgery. analysis. METHODS The analysis performed investigated Between October 1, 2007, relationships among type of closure, RESULTS and December 31, 2008, all location of tumor, patient age and sex, A total of 433 patient quespatients undergoing Mohs mianalgesia used, and level of tionnaires were returned and crographic surgery were given postoperative pain. included in the final analysis. a postoperative pain assessThe questionnaire was adment sheet upon completion The highest pain scores were associated ministered to 2197 patients, of surgery (Fig 1). Patients with the day of surgery, number of for a response rate of 19.7%. with contraindications to anlesions excised, flap and graft repair, For this type of nonincentivealgesic consumption (ie, liver younger age, and narcotics consumed. based questionnaire involvdisease or documented mediing several days of follow-up, cation allergy) were not inthe expected response rate was 10% to 15%.6 This cluded in this study. Before discharge, the patient response rate was satisfactory given the questionwas instructed to apply ice and take 500 mg of naire type and based on recent studies showing less acetaminophen orally (Extra Strength Tylenol, McNeil bias with lower questionnaire response rates.7,8 Consumer Healthcare, Fort Washington, PA) only Demographic criteria are included in Table I. In all, when needed for pain. Patients were not instructed 56% of patients were male and 44% were female. A to use analgesics prophylactically. The patients were total of 80% (n = 344) of patients had one lesion advised to refrain from using nonsteroidal anti-inflamtreated whereas 20% (n = 89) of patients had more matory drugs (NSAIDs) and aspirin to reduce postopthan one lesion treated on the day of surgery. The erative bleeding risk. If a patient requested prescription mean age of the patients was 67 years; the youngest pain medication, either 100 mg of propoxyphene treated was 21 years old, while the oldest was 100 napsylate and 650 mg of acetaminophen (Darvocet years old. As demonstrated in Table II, the average N100, Xanodyne Pharmaceuticals, Basking Ridge, NJ) size of the postoperative Mohs defect was 4.29 cm2. or 5 mg of hydrocodone and 500 mg of acetaminophen The median size was 2.7 cm2 with the smallest lesion (Vicodin, Abbott Laboratories, Parsippany, NJ) were at 0.14 cm2 and the largest at 67.5 cm2. prescribed. Table III demonstrates the frequencies of operaThe patients were instructed to rate their pain tive locations. The most common sites of surgery each morning and evening for 4 postoperative days. included the nose (30%), cheek (14.7%), forehead Their pain scores were self-rated using the Wong(9.9%), ear (9%), eyelid (5.3%), neck (5.1%), lip Baker FACES scale, a 0-to-10 scale using descriptive (4.4%), and temple (4.1%). In all, 82.5% of lesions faces to assess pain.3,4 In addition to the postoperative pain assessment, medication and dosage used undergoing Mohs micrographic surgery were on the for pain control, patient’s date of surgery and date of head and neck. birth, and the defect site, size, and type of closure for Table IV demonstrates the frequencies of the each lesion were recorded. repair type. The most common repair type was the Surgery sites were grouped into the following linear closure (69.5%). G&E was the second most locations: ear, nose, temple, cheek and preauricular, common type (13.4%). Certain patients with suforehead including eyebrow and hairline, lip includperficial tumors limited to the epidermis in chaling cutaneous lip, chin, eyelid including canthi, lenging locations were operated on using the scalp, neck, back, leg, arm, hand, foot, chest, and flexible scalpel technique.9 When the tumor was d

d

d

Firoz et al 81

J AM ACAD DERMATOL VOLUME 63, NUMBER 1

Fig 1. Postoperative pain assessment sheet.

Table I. Patient demographics Frequency

Male Female 1 Lesion [1 Lesion Age

Table III. Location of defect

Percent

242 191 344 89 N 433

56.0 44.0 79.5 20.5 Minimum Maximum Mean SD 21.00 100.00 66.97 12.56

Table II. Size of postoperative defect N

Length defect, cm 433 Width defect, cm 433 Defect size, cm2 433

Minimum Maximum Mean

0.40 0.20 0.14

9.00 7.50 67.50

SD

2.52 1.43 1.43 0.838 4.29 5.35

cleared using this technique, the final defect was allowed to heal by second intention. No reconstruction was performed for these patients. Grafts, both full and partial thickness, comprised 6.5% of the lesions. Flap repairs including advancement, rotation, and transposition flaps comprised 10.6% of repairs. Mean pain scores were tabulated for all patients. The mean pain scores for the day of surgery and postoperative days 1 through 4 are outlined in Fig 2. Mean pain scores were highest on the day of surgery, and successively decreased with each postoperative

Nose Cheek Forehead Ear Eyelid Neck Lip Temple Arm Scalp Leg Back Chest Hand Chin Abdomen Total

Frequency

Percent

130 64 43 39 23 22 19 18 16 16 15 9 8 6 4 1 433

30.0 14.7 9.9 9.0 5.3 5.1 4.4 4.1 3.7 3.7 3.5 2.1 1.8 1.4 .9 .2 100.0

day. Postoperative day 4 yielded the lowest mean pain score. Statistically significant differences in pain between postoperative days were demonstrated using the paired samples t test. Pain scores were significantly higher when comparing the day of surgery with postoperative day 1 (P \ .000), postoperative day 1 with day 2 (P \.000), postoperative day 2 with day 3 (P \.000), and postoperative day 3 with day 4 (P \ .000). There were no statistically significant

82 Firoz et al

J AM ACAD DERMATOL JULY 2010

Table IV. Type of repair Linear closure G&E Advancement FTSG Transposition STSG Rotation Island pedicle Total

Frequency

Percent

301 58 32 23 6 5 4 4 433

69.5 13.4 7.4 5.3 1.4 1.2 .9 .9 100.0

FTSG, Full-thickness skin graft; G&E, granulate and epithelialize; STSG, split-thickness skin graft.

Fig 3. Pain scores by postoperative day (POD) and medications consumed.

Fig 2. Pain scores by postoperative day. There is a statistically significant difference in pain scores between the day of surgery and postoperative days 1, 2, 3, and 4 (P = .000 for days 1, 2, 3, and 4 postoperatively). There is no significant difference in pain scores between AM (red ) and PM (black) on the same postoperative day (P = .973 day 1, P = .341 day 2, P = .894 day 3, P = .08 day 4).

differences in pain when comparing pain scores in the morning versus the evening on the same postoperative day (Fig 2). A more detailed graph of postoperative pain scores when accounting for type of medications consumed is provided in Fig 3. Patients who consumed narcotics had the highest postoperative pain scores, followed by patients who took acetaminophen, and finally patients who did not take any medication. Frequency and type of pain medications were tabulated across patients for each postoperative day (Fig 4 and Table V). Only 52% of patients took any pain medication on the day of surgery, and this number declined successively with each postoperative day. By postoperative day 4, only 5% of patients were taking medication for pain. Most patients took

Fig 4. Percent of pain medications consumed according to postoperative day.

500 mg of acetaminophen for pain control on the day of surgery, with only 7.1% taking any type of narcotic. The proportion taking narcotics declined to 4.1% on postoperative day 1, and to 0.6% by postoperative day 4. The mean pain scores on the day of surgery were compared between patients taking 500 mg of acetaminophen versus no medication using the independent samples t test to check for significance (Table VI). Patients taking no medication had significantly lower mean pain scores when compared with patients taking 500 mg of acetaminophen (P \.001). Patients taking narcotics had statistically significantly higher mean pain scores compared with patients taking 500 mg of acetaminophen (P \ .001) or patients taking no medications (P \.001). The relationship between size of the final Mohs defect and the types of medications consumed after surgery was also investigated. A one-way analysis of variance was performed, comparing the mean

Firoz et al 83

J AM ACAD DERMATOL VOLUME 63, NUMBER 1

Table V. Percent of patients taking medication Day of surgery

POD1

POD2

POD3

POD4

48.0 41.6 1.8 1.8 0 2.5 1.4 .5 .9 .7 .2

69.7 24.0 1.8 1.2 0 1.8 .2 .5 .2 .2 .2

82.2 14.3 1.2 0.5 0 1.2 .5 .2 0 0 0

89.4 7.4 .9 .2 .2 1.6 .2 0 0 0 0

94.5 3.9 .2 .2 0 .9 .2 0 0 0 0

No medication Acetaminophen Darvocet Vicodin Aspirin NSAIDs Hydrocodone Acetaminophen 1 Vicodin Acetaminophen 1 Darvocet Tramadol Acetaminophen 1 NSAID

NSAID, Nonsteroidal anti-inflammatory drug; POD, postoperative day.

Table VI. Comparison of mean pain scores by pain medication

No medication vs Acetaminophen Acetaminophen vs Narcotics No medication vs Narcotics

N

Mean

208 180 180 31 208 31

1.034 3.211 3.21 6.03 1.03 6.03

Mean differences

Significance (2-tailed)

e2.1775

\.001

e2.82

\.001

e5.00

\.001

differences in postoperative defect size among groups that consumed no medication, acetaminophen, or narcotics. There was no statistically significant relationship between defect size and type of medications consumed, (F{3,429} = 1.166, P = .322). In other words, patients using narcotics did not have larger postoperative defects when compared with patients using no pain medication or acetaminophen. The mean pain scores were also calculated for each type of closure on the day of surgery (Table VII). The highest mean pain scores were associated with full-thickness skin grafts and advancement, rotation, transposition, and island pedicle flaps. Patients healing with G&E (healing by second intention) had statistically significantly lower mean pain scores versus patients with flaps, linear closure, and grafts (Table VIII). These patients underwent only Mohs micrographic surgery without reconstruction. Patients with linear closures also had lower mean pain scores when compared with patients with flaps. Although this was not statistically significant (P = .052), it did show a trend toward less pain when comparing linear closures with flaps. The relationship between the type of pain medication consumed and the closure type was also examined. A cross-tabulation was performed and a

Table VII. Type of closure and pain Type of closure

Rotation Transposition Island pedicle FTSG Advancement Linear closure (CLC 1 ILC) G&E STSG

Mean pain score

5.00 4.67 3.50 2.70 2.63 2.47 1.03 0.80

CLC, Complex linear closure; FTSG, full-thickness skin graft; G&E, granulate and epithelialize; ILC, intermediate linear closure; STSG, split-thickness skin graft.

x2 analysis was performed for significance (Table IX). This analysis showed that the type of closure performed was significantly associated with the type of pain medication consumed (x 2{12, N = 433} = 46.15, Cramer V P = .000). A total of 3.3% of patients with linear closures and 1.7% of patients with no reconstruction (G&E) required narcotic prescriptions on the day of surgery compared with 17.4% of patients with flap reconstructions and 14.3% with full- and partial-thickness skin grafts. Patient age, patient sex, and number of lesions were also assessed for relationship to postoperative pain (Table X). A cut point of 66 years was chosen because this was the mean age of the study population. The mean pain score in patients younger than 66 years was higher when compared with patients older than 66 years (P \ .01). The mean pain score was not different in male versus female patients (P = .273). Patients with more than one lesion treated had significantly higher pain scores when compared with patients with only one lesion treated (P \.037). The location of surgery and its relationship to postoperative pain was also assessed (Table XI). The highest average pain score on the day of surgery was

84 Firoz et al

J AM ACAD DERMATOL JULY 2010

Table VIII. Comparison of mean pain scores by repair type and location Mean Mean differences

N

G&E vs 58 1.04 Flaps 46 3.17 G&E vs 58 1.04 Linear closure 301 2.46 G&E vs 58 1.04 Grafts 28 2.36 Flaps vs 46 3.17 Linear closure 301 2.46 Head and neck vs 362 2.41 Trunk and extremities 71 1.99

Significance (2-tailed)

e2.13

\.001

e1.42

\.0001

e1.32

.016

.715

.052

.420

.08

G&E, Granulate and epithelialize.

Table IX. Cross-tabulation of medications consumed on the day of surgery versus type of closure Type of closure

No medication Acetaminophen Narcotics NSAIDs Total

Linear

G&E

FTSG

STSG

Flaps

Total

138 135 18 10 301

43 13 1 1 58

5 14 4

4 1

20 18 8

23

5

46

210 181 31 11 433

FTSG, Full-thickness skin graft; G&E, granulate and epithelialize; NSAID, nonsteroidal anti-inflammatory drug; STSG, split-thickness skin graft. (x 2{12, N = 433} = 46.15, Cramer’s V P = .000).

for the lip, followed by the forehead, chest, leg, and nose. The postoperative locations were aggregated into two categories, namely the head and neck versus the trunk and extremities. These two groups did not exhibit any statistically significant differences in postoperative pain.

DISCUSSION This was a prospective study to assess the amount of postoperative pain our patients have after Mohs micrographic surgery in the setting of elective use of pain medication. In addition, we studied which features had an effect on the occurrence of this pain. Before the study, we hypothesized that the location, type of repair, and size of final defect would play a role in degree of pain after surgery. We hypothesized that sex and age would not be important for postoperative pain. We also conjectured that pain would be greatest in the day or two after surgery, and subside thereafter. We believed that the majority of our patients would not use medications to control their pain. To test these hypotheses,

we administered a postoperative pain questionnaire prospectively using a validated pain scale to all our patients. Interestingly, we found that the day of surgery had the highest associated pain scores. The pain subsided significantly by postoperative day 4. Approximately half of our patients took pain medication, and the majority taking pain medication used 500 mg of acetaminophen. It is our office policy to only prescribe narcotic pain medication at the patient’s request; we do not routinely prescribe narcotics to patients after surgery. For this reason, it is possible that the majority used acetaminophen because it was the only drug available. However, some patients requested prescription pain medication upon discharge from surgery because they were concerned about postoperative pain requiring stronger medication than acetaminophen. Other patients called and requested a stronger pain medication than acetaminophen on postoperative day 1 or 2, at which point a prescription was called in to the pharmacy. Those who took pain medication did have higher pain scores than patients who did not; this was true for patients taking prescription narcotic medication and over-the-counter pain medication. One potential limitation is that 3.4% of our respondents chose to take analgesics that we did not prescribe such as NSAIDs, tramadol, or aspirin. We still chose to include these patients because, although we did not prescribe them, these medications have been used safely as analgesics in other studies of postsurgical pain. We thought that surgery on the head and neck would be more painful than the trunk and extremities; this did not end up being true. We thought having more than one lesion operated on would be more painful, and this did show in the data analysis. We thought flaps would be more painful than linear closures or G&E, and this was confirmed. We were surprised that the size of the defect did not play a large role in pain or types of medications consumed; we initially thought the larger the defect size, the greater amount of pain. It was also interesting that the type of reconstruction did play a role in the type of pain medications consumedeflaps and grafts required more narcotic prescriptions than linear closures or G&E. The data show a predominance of linear closures, and this reflects the principal surgeon’s preference for this repair type; now he may add that it also yields less pain. Patient sex did not play a role in degree of pain, although younger patients did have higher pain scores. Several other studies have also investigated postoperative pain after surgery, although this is the only study to our knowledge that measured pain after Mohs micrographic surgery. A systematic review of

Firoz et al 85

J AM ACAD DERMATOL VOLUME 63, NUMBER 1

Table X. Patient demographics and pain Age $ 66 y Age \ 66 y Male Female 1 Lesion [1 Lesion

N

Mean

SD

SEM

223 206 241 191 343 89

2.06 2.64 2.23 2.48 2.22 2.80

2.18 2.42 2.17 2.47 2.25 2.48

0.146 0.168 0.140 0.179 0.122 0.263

Mean difference

95% CI Lower

95% CI Upper

Significance (2-tailed)

e0.58 e0.25

e1.02 e0.70

e0.140 0.198

.010 .273

e0.57

e1.11

e0.035

.037

CI, Confidence interval.

Table XI. Postoperative pain and location Location

Lip Forehead Chest Leg Nose Ear Chin Scalp Eyelid Temple Arm Cheek Neck Hand* Back Abdomen

Mean pain score, surgery day (PM)

N

3.32 2.86 2.86 2.73 2.69 2.54 2.50 2.38 2.04 1.89 1.88 1.80 1.36 1.00 0.67 0

19 43 7 15 130 39 4 16 23 18 16 64 22 6 9 1

*Only surgery location with greater pain on postoperative day 1 when compared with the day of surgery was the hand (1.833 mean pain score on postoperative day 1).

studies investigating the effect of preoperative, intraoperative, and postoperative analgesia on postoperative pain included NSAIDs, opioids, local anesthetic, and other analgesics.10 Surgery types included dental, abdominal, gynecologic, and minor orthopedic procedures. Visual analog scales for pain were used to measure significant differences in pain (P \ .05). They found that timing of analgesia, whether preoperative or postoperative, did not make a significant difference in pain scores. Another systematic review investigating preoperative NSAIDs (type II selective, or coxibs) on postoperative pain relief included 22 randomized trials.11 Analgesics used in this review included NSAIDs, coxibs, placebo, and acetaminophen before oral, abdominal, gynecologic, otolaryngologic, and orthopedic procedures. Twenty of these trials assessed postoperative pain and analgesic consumption within the first 24 hours. This review found that preoperative use of NSAIDs, when compared with placebo, showed reduced postoperative pain and

analgesic consumption, and reduced postoperative nausea and vomiting without an increase in intraoperative bleeding. Another interesting study investigated the use of the visual analog scale as a pain measurement tool in 150 patients on the day after intra-abdominal surgery.12 Patients rated their pain using the visual analog scale from 0 to 100 and the pain medications they consumed were also recorded. They found that grouping the pain scores into thirds predicted best the clinical relevance, or the clinical need for analgesia. They do state that this grouping may not have been appropriate if the pain scores were low overall. A study investigating cancer-related breakthrough pain during oral fentanyl therapy sought to evaluate clinically relevant changes in pain scores using a scale from 0 to 10. They found that converting the scales into percentages of pain improvement or changes in the pain scale of 2 points resulted in higher clinical relevance.13 A study that investigated the use of rofecoxib after total knee arthroplasty in 70 patients used the 0-to-10 visual analog scale to assess pain. The median pain scores between the placebo and treatment group were statistically different; 2.2 in the placebo group versus 3.5 in the treatment group during the inpatient stay, and 2.6 versus 3.7 at 1 week after discharge.14 Our study differed from the studies mentioned above in several ways. We did not administer analgesics before the surgery, nor did we prescribe NSAIDs or selective cox-II inhibitors. We chose to use a 10-point visual analog scale similar to other studies, although several studies used a 100-point visual analog scale. The ‘‘meaningful’’ difference in pain scale as it relates to clinical relevance is difficult to assess and compare with other studies because of the differences in type of surgery and patient characteristics. Our data reflected pain scores on the lower end of the scale when compared with other surgical procedures, possibly because of the less invasive nature of Mohs micrographic surgery performed in an outpatient setting and not requiring general anesthesia or intravenous sedation.

86 Firoz et al

J AM ACAD DERMATOL JULY 2010

There are limitations to the study. Not all patients returned the questionnaires, and perhaps those who returned the questionnaires had a higher overall level of pain. However, the average pain scores were on the lower end of the pain scale and did not reflect this. This cohort of patients only had Mohs micrographic surgery with defect reconstruction, and pain was not compared with patients undergoing other procedures. Finally, this was a private practice population in Houston, TX, which may be different when comparing results with large academic practices in other parts of the country. We describe postoperative pain patterns after Mohs micrographic surgery with defect reconstruction and show which factors produce more or less pain. We can thus appropriately alter our surgical decisions for the benefit of our patients, and better inform our patients about what to expect regarding postoperative pain. REFERENCES 1. About the joint commission. Available from: URL:http://www. jointcommission.org/AboutUs/. Accessed January 8, 2009. 2. Go¨ro¨cs TS, Lambert M, Rinne T, Krekler M, Modell S. Efficacy and tolerability of ready-to-use intravenous paracetamol solution as monotherapy or as an adjunct analgesic therapy for postoperative pain in patients undergoing elective ambulatory surgery: open, prospective study. Int J Clin Pract 2009; 63:112-20. 3. Wong D, Baker C. Pain in children: comparison of assessment scales. Pediatr Nurs 1988;14:9-17. 4. Stuppy DJ. The FACES pain scale: reliability and validity with mature adults. Appl Nurs Res 1998;11:84-9.

5. Morganroth PA, Gelfand JM, Jambusaria A, Margolis DJ, Miller CJ. A randomized, double-blind comparison of the total dose of 1% lidocaine with 1:100,000 epinephrine versus 0.5% lidocaine with 1:200,000 epinephrine required for effective local anesthesia during Mohs micrographic surgery for skin cancers. J Am Acad Dermatol 2009;60:444-52. 6. Survey response rates. Available from: URL:http://www.peo plepulse.com.au/Survey-Response-Rates.htm. Accessed June 2, 2009. 7. Keeter S, Kennedy C, Dimock M, Best J, Craighill P. Gauging the impact of growing nonresponse on estimates from a national RDD telephone survey. Public Opin Q 2006;70: 759-79. 8. American Association for Public Opinion Research. Response ratesean overview. Available from: URL:http://www.aapor. org/responseratesanoverview. Accessed June 2, 2009. 9. Vujevich JJ, Goldberg LH, Kimyai-Asadi A. Mohs micrographic surgery using a flexible blade for tumors of the scalp. Dermatol Surg 2009;35:1130-3. 10. Moiniche S, Kehlet H, Dahl JB. A qualitative and quantitative systematic review of preemptive analgesia for postoperative pain relief: the role of timing of analgesia. Anesthesiology 2002;96:725-41. 11. Straube S, Derry S, McQuay HJ, Moore RA. Effect of preoperative Cox-IIeselective NSAIDS (coxibs) on postoperative outcomes: a systematic review of randomized studies. Acta Anaesthesiol Scand 2005;49:601-13. 12. Bodian CA, Greedman G, Hossain S, Eisenkraft JB, Beilin Y. The visual analog scale for pain: clinical significance in postoperative patients. Anesthesiology 2001;95:1356-61. 13. Farrar JT, Portenoy RK, Berlin JA, Kinman JL, Strom BL. Defining the clinically important difference in pain outcome measures. Pain 2000;88:287-94. 14. Buvanendran A, Kroin JS, Tuman KJ, Lubenow TR, Elmofty D, Moric M, Rosenberg AG. Effects of perioperative administration of a selective cyclooxygenase 2 inhibitor on pain management and recovery of function after knee replacement. JAMA 2003;290:2411-7.