Development and validation of a practical score to predict pain after excisional hemorrhoidectomy
Francesco Selvaggi, MD, EBSQ colo; Gianluca Pellino, MD; Guido Sciaudone, MD, PhD; Giuseppe Candilio, MD; Silvestro Canonico, MD.
Unit of General Surgery, Second University of Naples, Naples, Italy
Running head: predictors of pain after hemorrhoidectomy
Please address correspondence to: Francesco Selvaggi, MD, EBSQ colo Associate Professor of General Surgery Unit of General Surgery Second University of Naples Via F. Giordani,42 80122 – Naples, Italy Mob. +39-3358419132
Fax: +39-815667919
e-mail:
[email protected]
1
ABSTRACT
Purpose: Excisional hemorrhoidectomy(EH) can be complicated by high early (EP) and prolonged pain (PP). This study aimed to determine predictors of high postoperative pain and to develop a risk score suitable to identify patients candidates to a more active analgesic treatment. Methods: We collected data of patients undergoing EH between January 2005 and September 2012(development group). Patient-,disease-, surgery-,and surgeon-related characteristics were gathered. Anxiety was evaluated by means of STAI-Y. EP was assessed at 1- while PP at 15-day follow-up by means of a 10cmVAS(cut-off:EP≥5 PP≥3cm). On the basis of the odds ratio from a logistic regression, independent risk factors were assigned a weighted integer. The sum comprised the risk score, which was validated on a prospective cohort of patients undergoing EH between September 2012 and January 2014. Results: 514 patients were included in development group. Incidence of EP was 18.3%,with 40.4% developing PP. Younger age, male gender, advanced education, constipation, external component, anxious state and trait, high anal resting tone were independent predictors of EP. Previous pain exerted a protective effect. Incomplete postoperative evacuation, advanced education level, high anal resting tone, and anxious trait were predictors of PP. In the development group, the risk of EP ranged between 1% (low-risk) and 21% (high-risk). ROC analysis of validation group(n=130) confirmed the discriminatory power of this model(area under the ROC =0.69). Conclusion: The score can stratify the risk of EP following EH, identifying high-risk patients candidates to active analgesic administration or alternative surgical procedures.
Key-words: postoperative complication; pain; hemorrhoids; hemorrhoidectomy; risk score 2
INTRODUCTION
Hemorrhoids (HD) are one of the most common anorectal disorders and III–IV degree HD are best treated with surgery [1-5]. HD is regarded as a minor procedure, but it can be associated with several postoperative complications, among which pain remains one of the most important patient complaints [2]. Pain represents the single most important reason why patients avoid surgery [3] or prefer stapled hemorrhoidopexy over excisional hemorrhoidectomy (EH), though higher rates of recurrences are being observed in the long-term [4,5]. It is still unclear why the duration/intensity of pain varies between subjects. Our aim was to determine factors associated with high early postoperative pain in patients undergoing EH, in order to develop and validate a risk score suitable to identify patients candidates to more active analgesic strategies or alternative procedures, eventually allowing prevention of such an invalidating condition.
3
MATERIALS AND METHODS
A development group was obtained by retrospectively reviewing data from a prospective database of patients undergoing EH for III- IV-degree HD between January 2005 and December 2012 in our Unit. Patient-, disease- and surgery-related issues were taken into account. EH consisted of Milligan & Morgan hemorrhoidectomy with either scissors and electric scalpel or LigaSureTM Precise [1]. Those who received additional maneuvers or procedures (mucopexy, circular hemorrhoidectomy [6]), were excluded from further evaluation (Figure 1). A validation group was prospectively enrolled between October 2012 and January 2014.
Preoperative assessment All patients underwent clinical examination and proctosigmoidoscopy or colonoscopy when indicated. Hemorrhoid grades were attributed following the definitions of the Standards Practice Task Force of The American Society of Colon and Rectal Surgeons[7]: grade III HD were defined as prolapse with Valsalva requiring manual reduction,
grade IV as
chronically prolapsed HD with ineffective manual reduction. HD-related symptoms (e.g. pain, bleeding) were thoroughly collected. All patients in the present study were submitted to anal manometry. The length of functional anal canal was evaluated with rapid pullthrough method. A “long functional anal canal” was defined by values > 4 cm in men and > 3 cm in women [8]. In our unit we routinely administer a questionnaire to assess the level of anxiety, the Y-State Trait Anxiety Inventory (STAI-Y) [9]. It is a validated and short questionnaire consisting of 2 items to assesses anxiety state (STAI-Y1) and trait (STAIY2). STAI-Y ≥50 was considered pathological.
4
Surgery All ≥18-year-old patients undergoing open EH were considered suitable for evaluation. The procedures were performed by 3 senior surgeons and by 5 residents under direct supervision of a Tutor. Results were analyzed according to the case load of operating surgeon: outcomes of Consultant surgeons (FS, SC, GS) were distinguished from those of surgeons in training. Most procedures were performed with either local anesthetic injection in outpatient settings or under spinal anesthesia, while only selected cases received general anesthesia. HD were infiltrated with an epinephrine and saline solution (1:200 000). Patients received intra-operative i.v. ceftazidime 2 g and metronidazole 1 g.
Postoperative management A 10-cm Visual Analogue Scale (VAS) was used to assess pain intensity on a scale of 0 (no pain) to 10 (unbearable pain) at post-operative day 1 (early pain, EP),and 15 (prolonged pain, PP). VAS ≥5 cm and ≥3 cm at 1- and 15-day follow-up respectively were considered abnormal. For PP also perioperative complications were taken into account. Analgesic treatment was delivered following the evidences reported in the literature [10,11], adjusted on the availability of drugs in our Country. Opioids were not routinely administrated postoperatively, these were only given if necessary to relief patients’ complaint. Conventional NSAIDs (ketorolac tromethamine, KT)+ paracetamol (PC) were administrated intra-operatively and postoperatively after 6 and 12 hours. In case of EP>6, administration of opioids was considered (tramadol). If patients still complained for pain after 6 hours, doses were incremented. Patients were discharged after one, 10 or 24 hours in outpatient, day-surgery or one-day-surgery settings respectively. Patients were sent home and given prescription for oral KT (50 mg three times a day) ± oral PC (1000 mg) for 3 days, lactulose (starting with 20 ml twice a day) and oral metronidazole. They were provided a mobile number to contact in case of unbearable pain or other complications. 5
Follow-up examinations were planned 2, 3 and 6 weeks after surgery in outpatient settings; at first follow-up visit or at discharge, we routinely ask patients to grade the completeness of first evacuation (complete vs incomplete/fragmented).
Statistical analysis and risk score development Development data set: Univariate predictors of EP were identified. These were included in a multivariable logistic regression to identify independent predictors of pain (entry level p
