Subcutaneous Morphine Pump for Postoperative Hemorrhoidectomy Pain Management Elsa T. Goldstein, M.D., Paul R. Williamson, M.D., Sergio W. Larach, M.D.

From the Department of Colon & Rectal Surgery, Orlando Regional Medical Center, Orlando, Florida

PURPOSE: Many anorectal procedures are currently being performed on an outpatient basis, hemorrhoidectomy being the exception because of the need for parenteral narcotics postoperatively. We investigated the effective- ness of a subcutaneous morphine pump (SQMP) for out- patient posthemorrhoidectomy pain control. METHODS: In Phase 1 of our study, 22 patients undergoing radical hemorrhoidectomy were started on an SQMP protocol postoperatively. Twenty-nine patients received conven- tional postoperative narcotic dosing. In Phase 2, 19 pa- tients enrolled in an SQMP protocol underwent hemor- rhoidectomy in an ambulatory setting. Length of hospi- talization, catheterization rate, and pain control were evaluated. RESULTS: In Phase 1, zero patients in the study group and two in the control group required additional hospitalization beyond 23 hours for pain control. The rates of catheterization were similar between the two groups. Pain control was considered satisfactory in 21 of 22 study patients. There was no correlation between pain level and morphine dose. Eighteen of 22 patients expe- rienced minor side effects, necessitating pump removal in two patients. In Phase 2, 18 of 19 patients on the SQMP were discharged from the recovery room. Cost analysis shows the combination of outpatient hemorrhoi- dectomy and the SQMP to be cost-effective in comparison with an inpatient stay. CONCLUSIONS: The SQMP enables hemorrhoidectomy to be done on an outpatient basis. It provides effective pain control, enjoys high patient ac- ceptance, and is cost-effective. [Key words: Patient-con- trolled analgesia; Hemorrhoids; Morphine; Pain; Postop- erative]

Goldstein ET, Williamson PR, Larach SW. Subcutaneous morphine pump for postoperative hemorrhoidectomy pain management. Dis Colon Rectum 1993;36:439-446.

M any anorectal p rocedu re s are current ly be ing p e r f o r m e d on an outpat ient basis. However ,

radical ( three-quadrant ) h e m o r r h o i d e c t o m y still necessi ta tes a per iod of hospi ta l izat ion pos topera- tively for parentera l pain control and poss ib le uri- nary catheterizat ion. Outpat ient hemor rho idec -

t omy would be feasible if a m e t h o d of pain control were available that was m o r e effective than cur- rent ly available oral narcotics. This s tudy was un- der taken to test the possibi l i ty of using a subcuta- neous m o r p h i n e p u m p (SQMP) in the outpat ient

Address reprint requests to Dr. Williamson: 110 W. Underwood Street, Orlando, Florida 32806.

439

sett ing for p o s t h e m o r r h o i d e c t o m y pain manage-

ment .

Pat ient-control led analgesia (PEA) has b e e n

gaining in popular i ty for pos topera t ive pain man-

agement . Convent ional PCA, however , requires in-

t ravenous (IV) access. More recently, subcutane-

ously admin i s t e red PCA has b e e n shown to be

equal ly efficacious. 1 It is more suitable for outpa-

t ient use since it e l iminates the need for IV access.

M E T H O D S

Phase 1

Twenty- two consecut ive pat ients unde rgo ing he-

m o r r h o i d e c t o m y by one su rgeon (P.R.W.) were

enro l led in the pi lot study. Owing to the inability

to predic t the effect iveness and compl ica t ions of

the SQMP, pat ients were admi t ted on the day of surgery for a 23-hour stay. The pro tocol cal led for

the p u m p to be started in the recovery room im- media te ly after surgery. However , because of con-

flicts with nursing protocol , the first 11 pat ients

had their p u m p started at the t ime of discharge

f rom the hospital. The subsequen t 11 pat ients had

their p u m p started in the hospital by the h o m e health nursing service.

The CADD-PCA | ambula to ry infusion pump,

Model 5800 (Pharmacia Deltec, Inc., St. Paul, MN),

was used. The p u m p can be p r o g r a m m e d to ad- minis ter both cont inuous and bolus doses with a lockout feature to p reven t overdosing. It contains a lockable 50-ml cassette for narcotics. The p u m p is l ightweight (15 ounces ) and small (1.1 inches x

3.5 inches x 6.4 inches) and is connec t ed by a 23- to 27-gauge butterf ly need l e to the subcu taneous tissue of the arm, trunk, or thigh. Morphine was used in a 5 -mg/ml concentra t ion. The need le site was covered with a t ransparent dress ing and ob- served daily.

Upon insertion, a bolus dose of 2 mg of mor- ph ine sulfate was given, fo l lowed by a cont inuous

440

Table 1. Concomitant Procedures

GOLDSTEIN ET AL

Procedure Study Control

Sphincterotomy 11 5 Fissurectomy - - 3 Colonoscopy ___ polypectomy 3 4 Sigmoidoscopy _+ polypectomy 2 6 Excision of anal papilla 2 - - Anoplasty 2 - - Correction of mucosal prolapse - - 2 Vasectomy - - 1 Rectocele repair - - 1 Hernia repair - - 1

infusion of 1 mg/hour for the first 24 hours. In addition, patient-controlled bolus doses of 0.5 mg were available every 15 minutes for three days.

Daily home visits were made by the nursing service to monitor pain relief and the development of side effects, with additional visits as necessary, i.e., for Foley catheter placement. Daily narcotic usage was recorded. The protocol could be altered to meet individual patient needs.

Patients were asked to rate their pain on a scale of 1 to 10 on postoperative days one, two, and three. They recorded the date and pain level of their first bowel movement after surgery. Side ef- fects of the morphine and the overall satisfaction with the pump were recorded.

The control group consisted of 29 consecutive patients undergoing hemorrhoidectomy by one surgeon (S.W.L.). They received intramuscular and

Table 2. Patient Characteristics

Study Control

Number of patients 22 29 Male 11 15 Female 11 14 Average age (range), in yr 42 (27-59) 46 (25-79)

Dis Colon Rectum, May 1993

oral pain medications as needed postoperatively. Both groups of patients underwent a radical (three- quadrant) hemorrhoidectomy using a closed tech- nique under monitored anesthetic control (i.e., sedation) with local anesthesia infiltration. Con- comitant procedures are listed in Table 1.

Phase 2

Nineteen consecutive patients undergoing he- morrhoidectomy in an ambulatory, setting were evaluated. The SQMP protocol was altered from that in Phase 1 to give a continuous dose of 1 mg/ hour for two days, with patient-controlled bolus doses of 1 mg available every 20 minutes. The pump was started in the recovery room, and pa- tients were discharged to their homes. Nursing visits were made as in Phase 1.

Data were analyzed using Student's t-test. A sta- tistically significant level was P = 0.05.

RESULTS

Phase 1

Demographic data of both the study and control groups reveal no statistically significant difference between the two groups; P = 0.093 (Table 2).

Length of Hospitalization. Two patients in the control group and none in the study group required hospitalization beyond 23 hours for pain control, not a statistically significant difference ( P = 0.217). Two patients in each group required a longer stay owing to the need for urinary catheterization (P = 0.778).

Incidence of Catheterization (Table 3). Ten pa- tients (45 percent) in the study group and seven (24 percent) in the control group required urinary catheterization. This difference was not statistically significant (P = 0.114). Within the study group, half had the pump started immediately postopera- tively, while the other half received traditional

Table 3. Patients Requiring Catheterization

Study Control

No. of patients 10 7 Male 4 4 Female 6 3 Age (mean) 40 43 Volume of urine 600 cc (200-1,000) 620 cc (300-1,000) Hours after operation 9.7 (6.33-20.33*) 11 (7.5-17.25)

* Voided spontaneously postoperatively; required a catheter the next day.

S U B C U T A N E O U S M O R P H I N E PUMP 441

22 Patients

intramuscular (IM) and p e r os (PO) pain medica-

tions until discharge, at which point the pump was placed. There was no statistically significant differ- ence in the catheterization rate between these two groups (P = 0.416) (Fig. 1).

Pain Leve l The average pain levels on a scale of 1 to 10 on days one, two, and three were 4.0, 3.8, and 3.7, respectively. Figure 2 demonstrates a more uniform pain level on day one, with a wider varia- tion on the following two days. Figure 3 shows that the first postoperative bowel movement occurs most commonly on day three. The pain level as- sociated with that bowel movement averaged 5.6, with a distribution most similar to the pain rating on day three (Fig. 4; compare with Fig. 2).

Pain Medications. On the first evening home,

# P t s lO 9 8 7 6

5 4

3 2 1 0

I I 11 Pump started 11 IM/PO pain in hospital meds post-op

- - 6 catheterized 4 catheterized

- - 5 voided 7 voided

Figure 1. Catheterization rate within the study group. There is no difference between the patients who were on the SQMP and those who were on conventionally admin- istered narcotics while in the hospital.

most patients required little more than the contin-

uous infnsion of morphine sulfate (Fig. 5). How-

ever, on the second and third days, the total daily dose varied considerably (Fig. 6). In four patients, the pump was used for four days, while one patient had it for six days.

While hospitalized, the control group took an average of 1.7 doses/day of ]M narcotics, in addi- tion to 2.1 doses/day of PO narcotics. In the study

group, 3 of 11 patients on the pump in the hospital required additional IM or PO narcotics for pain control.

Correlation B e t w e e n Doses a n d Pain Leve l

While on the pump, individual dose vs. pain level was evaluated. As can be seen in Figure 7, there was no correlation between these.

Side Effects (Table 4). Most patients reported minor side effects from the morphine, necessitat- ing pump removal in two. However, only one patient expressed d i s s a t i s f a c t i o n with the pump. There were no complications associated with the use of the pump.

Cost Analysis. The cost of the pump varies with the insurance companies' contracted price. It av-

erages $148.00 (range, $55-$265). Nursing visits average $72.50 (range, $55-$90 per visit). For three days, the average cost is $661.00 (range, $330- $1,065). A potential cost is that of a Foley catheter ($34).A 23-hour stay at the medical center for a

VOI. 36, No. 5

0 1 2 3 4 5 6 7 8 9 10 P a i n Level

Post-Op Day 1

Post-Op Day 3

Post-Op Day 2

Figure 2, Postoperative pain level in Phase 1.

442 GOLDSTEIN E T AL Dis Colon Rectum, May 1993

Figure 3. Day on which first bowel movement occurred postoperatively.

Figure 4. Pain level with first bowel movement.

hemorrhoidectomy costs $2,407, vs. $1,265 for the same procedure performed at a nearby free-stand- ing outpatient surgical facility. If the patient re- quires an additional hospital day for urinary cath- eterization, the hospital bill rises dramaticaliy to $6,594, as this is coded under the diagnosis-related group "excision of hemorrhoids with complica- tions." In Figures 8 and 9, the cost savings of

ambulatory surgery over a one-day stay are shown to be $481, while for a two-day stay the cost differ- ential is $4,634.

Phase 2

There were 19 patients (7 male and 12 female) with an average age of 47 years (range, 30-88 years).

#Pts 10

i

#Pts

8

6

4

2

0

Vol. 36, No. 5 SUBCUTANEOUS MORPHINE PUMP 443

Dose (mg) I I

C o n t i n u o u s ~ C o n t i n u o u s + . 5 - 2

Continuous +2.5-4 ~ Continuous +>10

Figure 5. Pain medication on the evening of postoperative day one.

10 9 8

7 6

5 4 3 2

1 0

0 2 4 6 8 !0 12 14 1~ 18 20 22 24 26 28 ~30 ~40

Morphine (mg)

Post-Op Day 2 ~ Post-Op Day 3

Figure 6. Pain medication on postoperative days two and three in Phase 1.

Catheterization. Four patients (21 percent), three male and one female, required urinary cath- eterization.

Length of Stay. Eighteen were discharged from the recovery room, while one patient stayed 23 hours because of nausea. No patient required read- mission.

Pain Level The pain level averaged 5.2, 4.1, and

4.7 on days one, two, and three, respectively, with a wide distribution among patients.

Pain Medications. The amount of pain medica- tion per 24 hours was higher in Phase 2 patients, owing to the change of dosage protocol. However, a wide variation among patients persisted (Fig. 10). The pump was used for four days in four patients.

Side Effects. Almost all patients experienced mi~

444 GOLDSTEIN ET AL Dis Colon Rectum, May 1993

Pain Level Figure 7. Morphine dose vs. pain level on postoperative day three.

Table 4. 48 Hour Stay $6594 Side Effects -Vs-

No. of Patients Reaction Outpatient surgery $1265 12 Drowsiness (freestanding outpatient 11 Itchiness facility)

2 Nausea 3 days SQMP 661 1 Rash 4 None Foley catheter 34

TOTAL $1960

23 hour stay $ 2407 - VS -

Freestanding outpatient 1265 facility

3 days SQMP 661

TOTAL $1926

Savings $ 481

Figure 8. Cost analysis of one-day stay vs. outpatient surgery.

nor side effects (Table 5), and one patient discon- t inued the pump. All patients expressed satisfac- tion with the pump.

Complications. There were no complicat ions re- lated to the pump.

D I S C U S S I O N

H e m o r r h o i d e c t o m y is associated with a dispro- port ionate amount of pain for the "size" of the

Savings $ 4634

Figure 9. Cost analysis of two-day stay vs. outpatient surgery (example).

operation. Because of this and the urinary re tent ion that f requent ly follows, it has been necessary to hospitalize patients postoperatively.

The PCA pump was first in t roduced in 1972. 2 Since then it has received widespread acceptance for postsurgical pain control . Its main advantage is that it allows the patient immedia te access to nar- cotics after h i s /her percep t ion of pain. Numerous studies have d o c u m e n t e d its efficacy, safety, and patient sat isfaction) 7 It has not been shown, how- ever, to alter the length of hospitalization or post- operative costs in abdominal surgery. < 8 The PCA pump delivers a loading dose, after which the patient self-administers bolus doses at preset inter- vals ( lockout) . These bolus doses are much smaller and are given more f requent ly than traditional par- enteral narcotics, thereby achieving a more con-

Vol. 36, No. 5 SUBCUTANEOUS MORPHINE PUMP 445

10

9 8

7 6

5

4

3

2

1 0

# P t s

I I I I 0 05 10 15 20

l z Z / /

Z z

I I I I I I 25 30 35 40 45 50

M o r p h i n e ( m g )

Post -Op Day 1

I r I 55 60 65

Post -Op Day 2

I I 70 75

Post -Op Day 3

80

Figure 10. Pain medication used by Phase 2 patients.

Table 5. Side Effects: Phase 2

No. of Patients Reaction

10 Drowsiness 9 Itchiness 8 Nausea/vomiting 3 Disorientation 2 Headache 1 Dizziness

stant drug level. 5' 9 This del ivery system should el iminate the wide swing in drug level seen in traditional IM dosing. 1~

A further modif icat ion of PCA allows a continu- ous infusion of narcotic to be s u p p l e m e n t e d by bolus doses as needed . This m e t h o d theoret ical ly should be more effective. Some studies have dis- proven any additional benefi t from the constant infusion) ' 1~ They showed that this me tho d in- creases the overall dose but does not decrease the number of boluses used or the pain level.

A more recent s tudy ~ evaluated IV v s . subcuta- neous PCA in patients undergo ing elect ive abdom- inal or extremity surgery. Although patients under the subcutaneous reg imen requi red higher total doses, there was no di f ference in overall satisfac- tion be tween the two groups. The advantage of the SQMP is that IV access is not required. This makes it suitable for outpat ient use.

Our study confirms the efficacy of the SQMP in achieving postoperat ive pain control. All but one patient in both study groups expressed satisfaction with the level of pain control achieved. The one dissatisfied patient was subsequent ly d iscovered to be a prescript ion drug abuser.

The rat e of urinary catheterizat ion is high in this study al though not out of line with recent re- ports lZ, 13 Perioperative fluid restriction has been shown to decrease the inc idence of urinary reten- tion in anorectal surgery. 13' 14 In this series, periop- erative fluid intake was not restricted. Additionally, overzealous nursing efforts resul ted in many pre- mature catheterizations that might have been

avoided had the patients not b een quer ied hour ly as to whether they had voided yet.

The SQMP did not alter the inc idence of urinary catheterization, which is somewhat surprising in that pain is felt to be a contr ibut ing factor leading up to it. 15 A recent report , ~6 however, ci ted two cases in which PCA use masked urinary retention, leading to increased discomfor t and morph ine con- sumption. Morphine was felt to inhibit the sensa- tion of urgency and to increase the sphincter tone, resulting in urinary retention.

The SQMP was not responsible for any postop- erative complicat ions in this study. One complica- tion that has b een descr ibed in the literature is respiratory depress ion occurr ing with excess se-

446

dation.5, 7, 1~ This should not be a p rob lem if the SQMP is correct ly administered. With narcotic an- algesics, analgesia occurs at a lower dose than does sedation. Clinically, this has been shown by the fact that patients using PCA have consis tent ly lower sedat ion and pain scores than those using conven- tional pain control methods . 6' 9 There were numer- ous minor side effects due to the morph ine and not related to the m o d e of delivery.

The variability seen in our patient 's pain levels and their narcotic requ i rements has been well de- scribed in comparable studies. ~' 2, 6 8 The lack of

correlat ion be tween pain levels and narcotic doses

was well demons t ra ted in this study. Prior studies, conduc ted in the setting of major

abdominal surgery, have not shown a decreased length of hospitalization or a cost d i f ference as- a result of the use of PCA. This study, however, clearly demons t ra ted the potential for major cost savings because of the fact that the hospitalization

could be reduced. No patient in the SQMP groups requi red hospitalization beyond 23 hours for pain management . Fur thermore , in Phase 2, 18 of 19 patients could be discharged from the recovery

room.

C O N C L U S I O N S

The SQMP provides a safe, effective, and eco- nomical means for outpat ient pos themorrho idec- tomy pain control, el iminating the need for inpa- tient parenteral analgesia. It does not alter the incidence of urinary retention. It enjoys a high rate of patient satisfaction despi te a significant num b er

of minor side effects.

A C K N O W L E D G M E N T S

The authors thank NMC Homecare , Longwood, Florida, for their assistance and participation in this study, Jane S. Gibson, Ph.D., for help with statistical analysis, and Michelle Oliver for graphic designs.

R E F E R E N C E S

1. Urquhart ML, Klapp K, White PF. Patient-controlled analgesia: a comparison of intravenous vs. subcuta- neous hydromorphone. Anesthesiology 1988;69: 428-32.

GOLDSTEIN E T AL Dis Colon Rectum, May 1993

2. Keeri4zanto M, Heaman S. Postoperative demand analgesia. Surg Gynecol Obstet I972;t34!647-51.

3. Parker RK, Holtmann B, White PF. Patient-controlled analgesia: does a concurrent opioid infusion im- prove pain management after surgery? JAMA 1991; 266:1947-52.

4. Albert JM, Talbott TM. Patient-controlled analgesia vs. conventional intramuscular analgesia following colon surgery. Dis Colon Rectum 1988;31:83-6.

5. Dubois MY. New technique in postoperative anal- gesia. Postgrad Med 1991;90:143-50.

6. Bennett RL, Batenhorst RL, Bivins BA, e t al. Patient- controlled analgesia--a new concept of postopera- tive pain relief. Ann Surg 1982;195:700-5.

7. Rayburn WF, Geranis BJ, Ramadei CA, Woods RE, Patil KD. Patient-controlled analgesia for post cesar- ean section pain. Obstet Gynecol 1988;72:136-9.

8. Rogers DA, Dingus D, StanfieldJ, DipiroJT, May JR, Bowden TA Jr. A prospective study of patient-con- trolled analgesia: impact on overall hospital course. Am Surg 1990;56:86-9.

9. Graves DA, Goster TS, Batenhorst RL, Bennett RL, Baumann TJ. Patient-controlled analgesia. Ann In- tern Med 1983;99:360-6.

10. Lutz LJ, Lamer TJ. Management of postoperative pain: review of current techniques and methods. Mayo Clin Proc 1990;65:584-96.

11. Vinik HR, Hammonds W, Lett A, Kissin I. Patient- controlled analgesia (PCA) combined with continu- ous infusion (CI). Anesth Analg 1990;70(suppl): $418.

12. Bleday R, Pena JP, Rothenberger DA, Goldberg SM, Buls JG. Symptomatic hemorrhoids: current inci- dence and complications of operative therapy. Dis Colon Rectum 1992;35:477-81.

13. Petros JG, Bradley TM. Factors influencing postop- erative urinary retention in patients undergoing sur- gery for benign anorectal disease. Am J Surg 1990;159:374 6.

14. Bailey HR, Ferguson JA. Prevention of urinary reten- tion by fluid restriction following anorectal opera- tions. Dis Colon Rectum 1976;19:250-2.

15. Corman ML. Colon and rectal surgery. Philadelphia: JB Lippincott, 1989:84-5.

16. Hodsman NB, Kenny GN, McArdle CS. Patient con- trolled analgesia and urinary retention. Br J Surg 1988;75:212.

17. Goudie TA, Allan MW, Lonsdale M, Burrow LM, Macrae WA, Grant IS. Continuous subcutaneous in- fusion of morphine for postoperative pain relief. Anesthesia 1985;40:1086-92.