Comparison of Postoperative Pain and Analgesia Requirement among Diabetic and Nondiabetic Patients undergoing Lower Limb Fracture Surgery - A Prospective Observational Study.

Background: Diabetic patients usually experience neuropathic pain and have a decreased response to opioids. Fractures are acute conditions and as such, they are very painful. No data is available related to fracture and postoperative pain in diabetics. Aim: This study was conducted to evaluate postoperative pain and analgesics requirement among diabetic and nondiabetic patients undergoing lower limb fracture surgery and the effect of glycosylated hemoglobin (HbA1c) on the postoperative pain. Setting and Design: This was a prospective observational study, conducted on 80 patients comprising of nondiabetic and diabetic, scheduled for elective lower limb fracture surgery under spinal anesthesia. Materials and
Methods: HbA1c was done in all the patients who were included in the study. Postoperative Visual Analog Scale (VAS) and analgesic consumption were assessed by an anesthesiologist blinded to the diabetic or nondiabetic status of the patients. VAS was assessed every 2nd hourly, for 24 h and rescue analgesia was given if the VAS was ≥4 and record was maintained. Sedation scores and adverse effects were also recorded postoperatively. Statistical Analysis: The Chi-square test was used for the analysis of categorical variables and Student's t-test was used for continuous variables.
Results: Diabetic group of patients had a significantly high VAS score with P ≤ 0.05. Rescue analgesics requirement was significantly different in two groups with diabetic patients requiring more supplementation of analgesia with a P = 0.025. The overall patient satisfaction was lesser in diabetic group (P = 0.004). There was statistically significant correlation between glycosylated hemoglobin and VAS at 2nd, 16th, 18th, 20th, 22nd, and 24th h.
Conclusion: Postoperative pain and analgesic requirement was significantly higher in diabetic patients with lower limb fracture. Glycosylated hemoglobin had good correlation with higher VAS. Copyright:

INTRODUCTION

Diabetes mellitus (DM) is characterized by alteration in carbohydrate metabolism, leading to hyperglycemia and increased perioperative morbidity and mortality. It evolves with diverse and progressive pathological changes. It has already been proved that diabetic patients experience neuropathic pain.[1] Diabetic patients with osteoarthritic knee joint have higher pain scores and opioid requirement in the postoperative period.[2] This has been attributed to diabetes-induced synovitis as they have higher interleukin 6 (IL 6) levels. Fractures are acute conditions and as such, they are very painful. The bone tissue itself does not contain nociceptors; bone fracture is painful for several reasons.[3] The inflammatory response starts with the fracture. Several mediators are released and will activate and sensitize primary sensory neurons; also, intense nerve sprouting that occurs in the fracture callus area is suggested to be involved in pain signaling. The hyperalgesia and allodynia after fracture indicate the development of peripheral and central sensitization; still, the underlying mechanisms are largely unknown.

Recent literature has proved the efficacy of glycosylated hemoglobin (HbA1c) for predicting outcomes in diabetic patients and also it reflects the degree of glycemic control in a very reliable manner.[45] Different studies have already proved that the postoperative opioid requirement is more in diabetic patients and those with higher preoperative HbA1c.[6] There is no literature on analgesic requirements on diabetic patients undergoing surgery for fixation of fractures. Hence, we planned this study to prove that there will be an increased demand of postoperative analgesics following lower limb fracture surgery in diabetics compared to nondiabetics and also aimed to evaluate the effect of HbA1c on postoperative analgesic requirements.

MATERIALS AND METHODS

After Institutional Research Ethics Committee (EC/NIMS/2215/2018) approval and after getting written informed consent for participation in the study and use of the patient data for research and educational purposes, the prospective observational cohort study was conducted from September 2018 to August 2019. The study included 80 patients aged between 18 and 70 years, of both genders and American Society of Anesthesiologists (ASA) physical status classes I to III, scheduled for elective lower limb fracture surgery done under spinal anesthesia. HbA1c was done in all the patients who were included in the study, after admission. Patients excluded were those with deranged coagulation profile, polytrauma patients, redo fracture surgery, complicated fracture or compound fracture, pathological fractures requiring combined spinal and epidural anesthesia, ASA PS class IV and more or patients with multiple comorbidities, patients with moderate to severe cardiovascular dysfunction, patients already on multiple analgesics due to diabetic neuropathic pain. DM is defined as a fasting blood sugar concentration ≥126 mg.dl−1 or the use of an oral hypoglycemic agent or insulin at the time of admission. Participants were divided into 2 groups with 40 patients in each group-Group “N“– nondiabetic patients and Group “D“-diabetic patients.

All the patients were explained the nature of the study and about their participation in evaluating postoperative analgesia. All the patients were familiarized with the Visual Analog Scale (VAS) in their native language. All the patients received oral pantoprazole 40 mg and alprazolam 0.25 mg, the night before surgery and in the morning 2 h before surgery, as premedication.

The preoperative monitoring included baseline heart rate (HR), electrocardiography (ECG), noninvasive blood pressure (NIBP), and oxygen saturation (SpO2). An intravenous (i.v.) access was established. Ringer lactate was used as the fluid. Under strict aseptic precautions, the subarachnoid block (SAB) was administered in the appropriate intervertebral space (L3–L4 or L4–L5) with Quincke 25 G spinal needle. After observing the free flow of cerebrospinal fluid, 2.5 cc of 0.5% heavy bupivacaine with 25 micrograms of fentanyl was administered. The sensory level was checked with the pinprick method and surgery started after attaining adequate level. Monitoring of vital parameters such as HR, ECG, NIBP, SpO2, and respiratory rate (RR) was done throughout the intraoperative period. Two segments sensory regression of spinal anesthesia was noted. Any additional analgesics such as i.v. paracetamol, ketorolac, tramadol, and fentanyl were avoided intraoperatively. Cases with an inadequate SAB, unanticipated technical difficulty in spinal, significant hypotension or high SAB requiring intubation, SAB wearing off early, prolonged surgeries requiring general anesthesia were discontinued from the study. The duration of surgery was recorded.

After the procedure, patients were shifted to the postanesthesia care unit/postoperative ward, where they were assessed second hourly, for 24 h, for the following parameters which includes VAS, rescue analgesic requirement, Wilson sedation scale, side effects, patient's overall satisfaction and the effect of glycemic control (HbA1c) on pain. Rescue medication/analgesics were given each time when the VAS was ≥4 and the choice of analgesic was left to the managing physician. We recorded the rescue analgesic requirement with ketorolac equivalent. Postoperative observation of parameters was done by the person who was blinded to the preoperative status of the patient or HbA1c values. Degree of pain assessed using the VAS of 0–10 with 0 = no pain and 10 = maximum intolerable pain.[7] Degree of sedation was assessed using the Wilson sedation scale as: -1 = fully awake and oriented, 2 = drowsy, 3 = eyes closed but arousable to command, 4 = eyes closed but arousable to mild physical stimulation (ear lobe tug), 5 = eyes closed and unarousable to mild physical stimulation.[8] The overall patient satisfaction with postoperative pain management was assessed using a Likert 5-point scale – 5 = excellent, 4 = good, 3 = fair, 2 = poor, 1 = bad.[9] Postoperative side effects/complications such as nausea, vomiting, shivering, pruritus, hypotension, bradycardia, respiratory depression (defined as RR <8/min or SpO2 <88%) and paresthesia were recorded.

Statistical analysis

The statistical software namely Statistical Package for the Social Sciences version 22 (IBM Corporation, USA) for Windows (Microsoft Corporation, USA). were used for the analysis of the data and Microsoft Word and Excel have been used to generate graphs, tables etc. Based on our pilot cases, mean VAS in the nondiabetic group being 4.7 ± 1.5 and in diabetic group 6.1 ± 2.1, the sample size of 38 in each group was calculated. To compensate for attrition, we selected a sample size of 100. Descriptive and inferential statistical analysis has been carried out in the present study. Results on continuous measurements are presented on mean ± standard deviation; minimum–maximum and results on categorical measurements are presented in number (%). Significance is assessed at 5% level of significance. Demographic data were compared using the t-test. The categorical variables were analyzed using the Chi-square test. The correlation coefficient was used to show the correlation of glucose control and HbA1c, with postoperative VAS. Leven's test for homogeneity of variance has been performed to assess the homogeneity of variance. To eliminate the confounding effect, the multivariate logistic regression analysis was done using rescue analgesic requirement as a dependent predictor and age, gender, ASA, body mass index (BMI), duration of surgery, and HbA1c as an independent predictors too find single a best predictor.

RESULTS

A total of 83 patients were enrolled in this study with analysis of 80 patients, 40 in each group as nondiabetic patients in Group N and diabetic patients in Group D. The flow of patients for the study is represented in Figure 1.

Figure 1

Flowchart of participants

Patients of Group N were younger with a mean age of 40.98 ± 15.58 and that of Group D were of higher age. Demographic distribution is given in Table 1. Most of the patients of Group N belonged to ASA PS class I (70%), followed by ASA PS class II (22.5%) and ASA PS class III (7.5%) and the major comorbidity in these patients was hypertension. Majority of the patients of Group D belonged to ASA PS class II (57.5%) [Table 1]. Almost in all the patients, with same drug dose and volume, level of block achieved was between T8 and T10. Two patients of Group D attained a higher level of a block of T4, which could be due to the patient factors such as short stature or obesity, or due to a difference in the patient positioning. Mean value of two-segment regression in Group D was lesser (76.50 ± 17.76) than that of Group N (80 ± 20.35) but was not statistically significant (P = −0.415) and the duration of surgery was significantly longer in Group D (108.88 ± 26.63, 126.88 ± 26.54, P = 0.003). The difference in HbA1c between the two groups was statistically significant (5.46 ± 0.32, 8.30 ± 0.89, P < 0.001). 25% of the patients in Group D had very high uncontrolled blood sugar levels with HbA1c more than 9% (with a significant difference between the two groups, the P < 0.001) [Figure 2].

Table 1

Demographic distribution in D and N groups

Demographic data	Mean±SD		P
	Nondiabetic group (n=40)	Diabetic group (n=40)
Age (years)	40.98±15.58	62.83±6.16	0.000**
Gender, n (%)
Males	28/40 (70)	20/40 (50)	0.068+
Females	12/40 (30)	20/40 (50)
Height (cm)	171.83±7.92	167.33±7.52	0.011**
Weight (kg)	67.23±10.52	71.28±10.05	0.082+
BMI (kg.m−2)	22.86±3.66	25.61±3.82	0.002**
ASA status, n (%)
Class I	28/40 (70)	0/40 (0)	0.000**
Class II	9/40 (22.5)	23/40 (57.5)
Class III	3/40 (7.5)	17/40 (42.5)
HbA1C	5.46±0.32	8.30±0.89	0.000**

+=Nonsignificant, **Statistically significant at the 0.01 level, *Statistically significant at the 0.05 level. HbA1c: Hemoglobin A1c, BMI: Body mass index, ASA: American Society of Anesthesiologists, SD: Standard deviation

Figure 2

Glycosylated hemoglobin percentage distribution in two groups of patients studied

VAS score was high in Group D over 24 h with a significant difference at 2nd, 4th, 8th, 10th, 18th, 20th, 22nd, and 24th h postoperatively with P ≤ 0.05 [Figure 3]. Wilson Sedation Scale showed a statistically significant difference between Group N and Group D in the 14th (2.45 ± 0.81, 2.88 ± 0.46, P = 0.005) and 16th (2.23 ± 0.97, 2.85 ± 0.75, P = 0.001) hours respectively with Group D having higher sedation score. Rescue analgesics requirement was significantly different in two groups with diabetic patients requiring more supplementation of analgesia with a P = 0.025. The analgesic requirement was categorized with ketorolac equivalent and found to have higher need of ketorolac equivalents in Group D (P < 0.001). As matched cohorts were not taken and age, BMI, duration of surgery, gender, ASA along with HbA1c were significant independent predictors of analgesic requirement measured in terms of ketorolac equivalents by independent sample t-test, Further multivariate logistic regression analysis was done using age, gender, ASA status, BMI, duration of surgery and HbA1c with analgesic requirement and Hosmer-Lemeshow test applied to get HbA1c ≥7 as single best predictor of analgesic requirement in given sample (P < 0.001).

Figure 3

Difference in Visual Analog Scale in two groups shown by bar charts with error bars

Common side effects observed in this study population were nausea, vomiting, shivering, hypotension and paresthesia. A statistically significant difference was observed in hypotension between the two groups with Group D showing a higher incidence (27.5% in Group D and 5% in Group N, P < 0.05). The incidence of shivering was higher in Group N but not statistically significant (P = 0.172). One patient of Group N complained of paresthesia in the postoperative period. Postoperative nausea and vomiting may be due to opioid usage [Figure 4].

Figure 4

Side effects observed in two groups

Overall patient satisfaction of Group N was better as 25 out of 40 patients expressed that they had an overall “good“ experience with the perioperative pain management and in Group D, a greater number of patients (28 out of 40 patients) said that they had a “fair“ experience with pain management. P = 0.004 was statistically significant. As HbA1c value was more than 6 in D group, we correlated HbA1c with VAS score in diabetic group and statistically significant correlation was observed at 2nd, 16th, 18th, 20th, 22nd, and 24th h, as represented in Table 2.

Table 2

Correlation of hemoglobin A1c with Visual Analog Scale in diabetic patients (n=40)

VAS (h)	Correlation coefficient (r)	P
2nd	0.551	0.000**
4th	0.280	0.080+
6th	0.096	0.554
8th	0.141	0.385
10th	0.287	0.073+
12th	0.290	0.070+
14th	0.049	0.766
16th	0.392	0.012**
18th	0.317	0.046*
20th	0.429	0.006**
22nd	0.506	0.001**
24th	0.457	0.003**

+Nonsignificant correlation, *Correlation is significant at the 0.05 level (two-tailed), **Correlation is significant at the 0.01 level (two-tailed). VAS: Visual Analog Scale

DISCUSSION

Acute postoperative pain is experienced by most of the patients undergoing surgery. Acute pain can turn into chronic pain if inadequately treated leading to negative psychological effects including anxiety. Some of the determinants of persistent postoperative pain include preexisting pain, genetic predisposition, psychological factors (e.g.: Preoperative anxiety and depression), female sex, obesity, smoking, younger age and duration of surgery.[10] Approximately a quarter of the patients who underwent inpatient orthopedic surgery, experienced postoperative pain even at 3–12 months postdischarge, which adversely affected their daily living.[11] Therefore, effective anesthetic/analgesic measures are important in the perioperative period to prevent progression to persistent pain.[10] Diabetes has been found as one of the most important factors leading to more requirements of analgesics in the postoperative period and increased complications in orthopedic surgery, especially if uncontrolled.[612131415] Studies have proved more analgesic consumption in diabetics postoperatively.[614] Even osteoarthritis and diabetes combined led to more analgesic consumption.[2] In our study, VAS scores were higher in patients of Group D than in patients of Group N. Rescue analgesic requirement was more and statistically significant in the diabetic group than in nondiabetics (P < 0.001). In a study by Wienemann et al., they concluded that physiological nociception and posttraumatic hyperalgesia to pressure is diminished at the foot with severe painless (diabetic) neuropathy.[16] This suggests that patients with painless diabetic neuropathy experience less posttraumatic/postsurgical pain in comparison to patients with painful diabetic neuropathy. This may be due to diabetic nerve damage leading to less pain. Similar effects cannot be postulated on fracture pain or postsurgical pain. In our study, it was observed that the timing of first postoperative rescue analgesic requirement was significantly delayed in Group N as they had lesser pain scores and the total dose and frequency of postoperative analgesics requirement was also lesser in them. Previous studies have proved that female sex, obesity and younger age, are associated with more postoperative pain.[1017] Our study results are similar to this except that in D group; we had more of elderly patients. Hence, uncontrolled diabetes may be more contributing to pain than age. We observed a significant correlation between HbA1c values and postoperative VAS scores, within the diabetic population, mainly in the 2nd, 16th, 20th, 22nd, and 24th postoperative hours. This is in concordance with the previous study where in 52 diabetic patients undergoing open nephrectomy we re-evaluated for the effect of long-term glucose control on postoperative analgesia requirements and concluded that, in diabetic patients, the preoperative HbA1c was associated with the postoperative fentanyl consumption. A significant correlation between preoperative HbA1c concentration and total fentanyl use during the first 48 h after surgery (r = 0.455, P = 0.001) was observed.[14] However, there are some studies which have conflicting results. One animal study demonstrated that systemic administration of streptozotocin induces mechanical hyper nociception that does not depend on hyperglycemia.[18] Furthermore, another study was done by Morley et al. titled “Effect of glucose on pain perception in humans,“ in which they demonstrated that the pain threshold was lower in humans who were rendered acutely hyperglycemic due to glucose injection, but in diabetic patients, they could not demonstrate a significant change in pain threshold.[19] Numerous studies have proved that diabetes decreases the pain threshold leading to more analgesic requirements. Our study showed similar results and also, we could observe a significant positive correlation between glycosylated hemoglobin and VAS. Higher sedation score despite higher VAS could be due to this group of patients receiving more analgesics.

There are few limitations in this study; the VAS was used to assess pain. There could be a variation in pain experienced by the patients in this study as it included all the lower limb fractures such as hip fractures, shaft femur fractures, and patella, tibia and ankle fractures. Another limitation of this study is omission of preoperative pain levels in the patients.

CONCLUSION

Glycosylated hemoglobin had good correlation with higher postoperative pain and higher analgesic requirement, suggesting significant-high pain scores in patients with uncontrolled sugars.

Highlights

DM evolves with diverse and progressive physiological changes. It has already been proved that diabetic patients experience neuropathic pain

Fractures are acute conditions and as such they are very painful. The inflammatory response starts with fracture. Hence, fracture in diabetics needs to experience more pain than normal and nondiabetic patients

Hence, we planned this study to prove that fracture pain will be aggravated in diabetics

Our study found that the postoperative pain and analgesic requirement were significantly higher in diabetic patients posted for lower limb fracture surgery. Glycosylated hemoglobin had good correlation with higher VAS and higher analgesic requirement, suggesting significant-high pain scores in patients with uncontrolled sugars

This guides us to be more attentive for pain management in patients with diabetes, especially uncontrolled diabetes to avoid its chronic complications and psychological impact as this population is more sensitive.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.