Patients presenting to four burn centers (Jaycee Burn Center, University of North Carolina, Chapel Hill, NC; The Burn Center, Washington Hospital Center, Washington, DC; Wake Forest University Baptist Burn Center, Wake Forest, NC; Nathan Speare Regional Burn Treatment Center, Upland, PA) within 72 hours of thermal burn injury between June 2009 to January 2011 were evaluated for study eligibility. Patients who were clinically unstable or who had coincident non-burn injury were excluded, as were prisoners, pregnant patients, patients with intentional injury, patients with a psychotic disorder, non-English speaking patients, and patients with hepatic failure, renal failure, or a history of chronic opioid use (defined as ≥ 20mg/day of oxycodone or equivalent). Also, because patients in this observational study were subsequently to be evaluated for participation in a randomized controlled medication trial, patients with greater than 20% total body surface area (TBSA) burn, patients with an estimated hospital stay of <5 days or > 40 days, patients with greater than first-degree atrioventricular block, patients taking a β-adrenergic antagonist medication, and patients with asthma, diabetes, coronary artery disease, and congestive heart failure were also excluded. Patients were also excluded who in the opinion of the investigators would not provide reliable data. Local Institutional Review Board approval was obtained from all study site IRBs.
Eligible patients were approached by research staff for study participation within 48 hours of burn center admission. Written informed consent was obtained from all participants. Study participation included blood sample collection for genetic analysis and completion of daily pain symptom interviews on two consecutive days following enrollment. During each daily pain symptom interview, participants were asked to rate their worst pain, their least pain, and their average pain over the past 24-hours, as well as their pain upon waking. Each pain assessment was performed using a verbal 0-10 numeric rating scale (NRS), where ‘0’ was defined to the patient as ‘no pain’ and ‘10’; as ‘pain as severe as it could possibly be’. A verbal NRS was used because it has been validated as a substitute for the Visual Analogue Scale in acute care settings,17
and because it does not rely on upper extremity use to make a precise mark on a scale. Daily pain symptom interviews which included the above assessments were performed on two consecutive days, beginning on the day of enrollment.
For study analyses, our initial intention was to combine the two ratings of daily average pain for each patient (one rating each day obtained on two consecutive days). However, as daily pain symptom interviews were being conducted, it was observed that patients sometimes provided an “average” pain rating greater than their worst reported pain or less than their least reported pain. These observations, together with the high degree of educational disadvantage in this population, lead us to appreciate that our study question (“Please rate your average pain during the last 24 hours…”) was poorly designed for the study population and (due to misunderstanding) did not yield valid data. Because of this, instead of using average pain, we took advantage of the multiple different pain assessments available within each individual (waking, worst, and least pain on two consecutive days) and created a measure of overall pain burden using linear mixed modeling described below. This measure of overall pain was used as the primary outcome measure for all analyses. Mean scores for waking, worst, and least pain over two days were also obtained from this model.
Study participants received analgesics as per standard study site burn care; no changes were made to the pain treatment of study participants. Information regarding medications received during the two-day study period was extracted from the medical record. For each opioid analgesic medication, total dose received during the two-day study period was calculated and then multiplied by a conversion factor referenced to a 30 mg dose of morphine.18
These doses were then summed to provide the total opioid dose (in morphine equivalents) received by the patient during the two-day study period. Benzodiazepine conversions were similarly calculated using a 10 mg diazepam reference.19
Demographic information was obtained during the initial patient assessment via standardized questionnaire. Information regarding patient burn characteristics was obtained from the medical record.
Blood samples were obtained for genetic analysis using an EDTA Vacutainer collection tube (BD, Franklin Lakes, New Jersey, USA). DNA was purified from whole blood samples using the QIAamp DNA Mini Kit (QIAGEN, Valencia, California, USA) on the QIAcube (QIAGEN), as per manufacturer's instructions. Genotyping was performed using a TaqMan Allelic Discrimination Assay for rs4818 on the Bio-Rad CFX96 Real-time PCR Detection System (Bio-Rad, Hercules, California, USA) at either the University of North Carolina (Chapel Hill, NC) or Washington Hospital Center (Washington, DC). Patient DNA samples were genotyped together with six HapMap CEU DNA samples (two of each genotype) and two “no template” control samples.
When multiple disease susceptibility variants occur in the same gene, the overall functional state of the gene may not be easily deduced from information regarding a single nucleotide polymorphism (SNP).20
For this reason we used a haplotype-based approach to examining COMT
variants. In a previous study, three haplotypes located in the central COMT
locus accounted for approximately 96% of all haplotypes in this region and were associated with variations in pain sensitivity15
and post traumatic pain.16
One of these haplotypes, the “low pain sensitivity haplotype”, codes for high COMT enzyme activity and is associated with relatively low pain vulnerability (i.e. is protective against pain).15
As in a previous study,16
we defined patients with no copies of this low pain sensitivity haplotype as having a “COMT
pain sensitive genotype”. Patients with a COMT
pain sensitive genotype were identified by genotyping SNP rs4818, because approximately 95% of individuals with a CC genotype at rs4818 have a COMT
pain sensitive genotype.15
Statistical analyses used linear mixed modeling to evaluate the association between COMT
pain vulnerable genotype and pain outcomes. Six pain measurements (waking, worst, and least pain for day 1 and day 2) for each individual were entered into the model as a correlated outcome variable. The correlations between pain measurements within each individual were taken into account by specifying nested random effects for intercept. The measure of overall pain was obtained as an adjusted least square mean pain score incorporating waking, worst, and least pain for day 1 and day 2. Mean scores for waking, worst, and least pain over two days were also obtained from this model. Age, gender, TBSA, burn depth, and time from admission to pain assessment were considered as important covariates and included in the model. In addition, because the frequency of genetic variations can vary by ethnicity,21
associations between COMT
pain vulnerable genotype and pain outcomes were also adjusted for patient ethnicity (European American vs. African American/Other). All analyses were conducted using SAS (version 9.2, SAS Institute Inc., Cary, NC). P-values < .05 were defined as statistically significant.