Systematic review doi:10.1111/j.1463-1318.2008.01638.x

Stapled haemorrhoidopexy for the treatment of haemorrhoids:a systematic review

J. Burch*, D. Epstein†, A. Baba-Akbari Sari*, H. Weatherly†, D. Jayne‡, D. Fox* and N. Woolacott*

*Centre for Reviews and Dissemination, †Centre for Health Economics, University of York, York and ‡St James�s University Hospital, Leeds, UK

Received 17 April 2008; accepted 30 April 2008

Abstract

Objective This systematic review aimed to evaluate the

short- and long-term safety, efficacy and costs of stapled

haemorrhoidopexy (SH) compared with conventional

haemorrhoidectomy.

Method We searched 26 electronic databases and web-

sites for studies in any language up to July 2006.

Inclusion criteria were predefined, and each stage of the

review process was conducted in duplicate.

Results Twenty-seven randomized controlled trials were

included (n = 2279). All had some methodological flaws.

Postoperatively, 19 trials (95%) reported less pain, 17

(89%) reported a shorter operating time, 14 (88%) a

shorter hospital stay, and 14 (93%) a shorter convales-

cence time following SH. However, prolapse was signif-

icantly more common after SH (OR 3.38; 95% CI: 1.00,

11.47). In the longer term, prolapse was significantly

more common after SH (OR 4.34; 95% CI: 1.67, 11.28)

as was reintervention for prolapse (OR 6.78; 95% CI:

2.00, 23.00). There were no differences in the rate or

type of complications. Conventional haemorrhoidectomy

and SH had similar costs during the initial admission.

Conclusion Compared with conventional haemorrhoi-

dectomy, SH resulted in less postoperative pain, shorter

operating time, a shorter hospital stay, and a shorter

convalescence, but a higher rate of prolapse and reinter-

vention for prolapse.

Keywords Haemorrhoidectomy, PPH, circular stapler,

Milligan–Morgan, Ferguson

Introduction

Haemorrhoids arise from engorged venous plexuses of

the anal canal, and can cause prolapse, bleeding, pain and

pruritis [1–5]. Symptomatic haemorrhoids are classified:

grade I: symptomatic, not prolapsed; grade II: prolapsed,

reduce spontaneously; grade III: prolapsed, require

manual reduction or grade IV: permanently prolapsed

[3,6]. Grades III and IV, and refractory grade II

haemorrhoids, require surgery [1,3]. Conventional

haemorrhoidectomy involves submucosal excision of the

prolapsing haemorrhoidal pedicles. Anodermal wounds

are either left open (Milligan–Morgan) or closed

(Ferguson). Approximately 8000 haemorrhoidectomies

were performed in England in 2004 ⁄ 5 [7].

In 1998, Longo introduced stapled haemor-

rhoidopexy (SH), which involves simultaneous

circumferential haemorrhoidal excision and mucosal anas-

tomosis, purportedly removing excess haemorrhoidal

tissue and returning the residual haemorrhoidal mass to

its original position [8,9]. Approximately 1500 SHs are

thought to have been conducted in the UK between 1998

and 2002 [10]. Alleged reductions in postoperative pain,

hospital stay and time to normal function are offset against

increased operative costs and perceived increases in

prolapse and postoperative complications [2,11–15].

The current review evaluates the short- and long-term

safety, efficacy and costs of SH in comparison to conven-

tional excisional haemorrhoidectomy (CH).

Method

Search strategy and selection criteria

We searched 15 electronic databases and seven resources

for clinical guidelines and systematic reviews (to July

Correspondence to: Dr Jane Burch, Centre for Reviews and Dissemination (CRD),

University of York, York YO10 5DD, UK.

E-mail: jb67@york.ac.uk

This report was commissioned by the NHS R&D HTA Programme as project

number HTA 05 ⁄ 21; the views expressed in this report are those of the

authors and not necessarily those of the NHS R&D HTA Programme. The

funding body had no role in the study selection, data extraction, the analysis and

interpretation of data, the writing of the report, or the decision to submit the

paper for publication. The corresponding author had full access to all the data

used in the review.

Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244 233

2006) without language restrictions, bibliographies of

included studies and relevant reviews, websites of five

relevant organizations and the content pages of five key

journals (July 2005 to July 2006). The full search strategy

is available elsewhere [16].

Randomized controlled trials (RCTs) with 20 or

more participants of any age, with prolapsing haemor-

rhoids, comparing circular SH with any CH technique

using scalpel, scissors or diathermy for excision were

eligible for inclusion. Trials of linear staplers, circular

staplers not designed for SH or patients undergoing

emergency procedures for thrombosed haemorrhoids,

were excluded. Results for postoperative (within

6 weeks) and long-term (12 months and beyond) out-

comes are presented. Full results are available elsewhere

[16].

Review process

Two reviewers independently screened studies for rele-

vance. Abstracts were included if sufficient outcome data

were available and the authors were contacted. Data

extraction and quality assessment were conducted by

one reviewer and independently checked by a second.

Quality was assessed using standard checklists adapted to

incorporate topic-specific issues [17]. Disagreements

were resolved through consensus or referral to a third

reviewer.

Data synthesis

Odds ratios (OR; dichotomous outcomes) or mean

differences (continuous outcomes), along with 95%

confidence intervals (CI) were calculated. Pooled OR or

weighted mean differences, and 95% CI were calculated

where no statistically significant heterogeneity was

observed. A random effects model was used unless there

were three or less studies, where a fixed effect model was

used. Meta-analyses were conducted in REVMAN 4.2.9

(Cochrane Collaboration). Statistical heterogeneity was

assessed using the v2 test and I2 statistic. Sources of

heterogeneity (degree of haemorrhoids; stapling gun;

comparator; age; sample size; date and country of trial;

anaesthesia; co-morbidity; study quality) and loss to

follow-up were investigated using sensitivity analyses. The

results were reported when they altered those of the

original meta-analysis.

The mean visual analogue scale (VAS) score for pain

across all the trials, at each time point, for each treatment

was estimated using Bayesian meta regression (WINBUGS,

version 1.4) [18]. The mean VAS score and the difference

between the treatments were assumed to decline linearly

from time from surgery. Statistical models where the VAS

score and the treatment effect were allowed to be

nonlinear were tested, but did not fit the data as well as

the linear model. These results are reported elsewhere

[16].

Costs of the procedures

Mean costs of the initial hospital admissions were

estimated for SH and CH at 2005 ⁄ 06 prices. Mean

operating time and length of stay were obtained from the

clinical review. The unit cost of a day in hospital was

obtained from national published sources [19] and the

list price of surgical costs from the manufacturers [20].

Costs of other hospital resources were assumed to be the

same for both procedures.

Results

The searches identified 653 references. Of these 147 full

papers were retrieved. Twenty-seven RCTs met the

inclusion criteria (n = 1137 SH; n = 1142 CH). The

flow of studies through the review is shown in Fig. 1, and

brief study details in Table 1. Full details are available

elsewhere [16].

653 identified

147 full papersretrieved

45 Background 9 Systematic reviews 3 Economic evaluations

506 Irrelevant

90 14 Case reports/series

58

43

15 Evaluated staple gun not designed for SH

36

7 Protocol/abstracts: insufficient information

Included clinical studies27 RCTs

(35 publications: 2 long-term follow-up and 6 duplicate publications)

1 Outcomes of interest not reported

76 18 NonRCTs

Figure 1 Flow of studies through the review process.

Stapled haemorrhoidopexy J. Burch et al.

234 Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244

Table 1 Details of included studies.

Study trial dates

n

SH ⁄ CH

Age

(as reported)

Male

(%)

Grades included

(as reported)

Staple gun

Comparator CH

Ascanelli (2005) [33]

2001–2003

50 ⁄ 50 Range: 30–73 21% II + III Mechanical suture

M&M + diathermy

Basdanis (2005) [36]

2000–2002

50 ⁄ 45 Range: 22–72 57% III: 73

IV: 22

PPH 01

M&M + diathermy

and ligasure

Bikhchandani (2005) [31]

2001–2003

42 ⁄ 42 Mean: 47 (NR) 83% III: 71

IV: 13

PPH 01

M&M

Boccasanta (2001) [25]

1996–1999

40 ⁄ 40 Mean: 51 (21–92) 41% IV: 80 PPH 01

M&M + HLB

Cheetham (2003) [14]

Trial dates: NR

15 ⁄ 16 Range: 26–72 71% Prolapsing: 31 PPH 01

M&M + diathermy

Chung (2005) [41]

2001–2003

43 ⁄ 45 Mean: 45.7 (NR) 67% III: 88 PPH 01

M&M + harmonic

scalpel

Correa-Rovelo (2002) [30]

Trial dates: NR

42 ⁄ 42 Mean: 45.15 (27–77) 49% III: 60

IV 24

Unspecified

stapling gun

Ferguson

Docherty (2001) [42]

Trial dates: NR

26 ⁄ 20 NR NR NR Unspecified

stapling gun

Ferguson

Gravie (2005) [46]

1999–2000

63 ⁄ 63 Mean: 47.5 (NR) NR Reducable: 85%

Nonreducible: 5%

No prolapse: 2%

PPH 01

M&M

Hasse (2004) [44]

1998–2001

40 ⁄ 40 Mean: 47.1 (NR) 49% III: 80 PPH 01

Fransler and

Anderson

Hetzer (2002) [39]

1999–2000

20 ⁄ 20 Mean: 47.6 (28–74) 73% II: 12

III: 28

PPH 01

Ferguson

Ho (2000) [28,45]

1999–2000

57 ⁄ 62 Mean: 48.6 (NR) 50% II + III PPH 01

M&M + diathermy

Kairaluoma (2003) [26]

1999–2000

30 ⁄ 30 Range: 17–65 53% III: 60 PPH 01

M&M + diathermy

Kraemer (2005) [27]

Trial dates: NR

25 ⁄ 25 Range: 28–82 54% III: 46

IV: 4

PPH 01

M&M + ligasure ⁄Fransler-Arnold

Krska (2003) [37]

Trial dates: NR

25 ⁄ 25 Mean: 50.8 (NR) 74% III: 50 Unspecified

stapling gun

M&M

Lau (2004) [22]

2001–2002

13 ⁄ 11 Mean: 49.1 (NR) 46% II: 13

III: 6

IV: 4

PPH 01

M&M + diathermy

Ortiz (2002) [43]

1999–2000

27 ⁄ 28 Mean: 47.6 (NR) 58% III: 29

IV: 26

PPH 01

M&M + diathermy

Ortiz (2005) [24]

2001–2002

15 ⁄ 16 Mean: 48 (28–69) 61% IV: 31 PPH 01

M&M + diathermy

Palimento (2003) [38,47]

1999–2000

37 ⁄ 37 Range: 25–84 64% III: 34

IV: 40

PPH 01

M&M + diathermy

Pavlidis (2002) [21]

1999–2000

40 ⁄ 40 Mean: 47.5 (29–75) 59% II: 16

III: 55

IV: 9

PPH 01

M&M + diathermy

Ren (2002) [32]

Trial dates: NR

45 ⁄ 45 Range: 29–82 67% III: 68

IV: 22

PPH 01

M&M

J. Burch et al. Stapled haemorrhoidopexy

Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244 235

Study quality

All included RCTs had some methodological flaws. Only

37% reported appropriate methods of randomization

and ⁄ or allocation concealment, 4% reported blinding

patients, 19% blinded outcomes assessors and 7% had

> 80% loss to follow-up at the final time point. Three

trials recruited a representative patient spectrum (grades

II, II and IV haemorrhoids), but either did not report the

method of randomization [21,22], allocation conceal-

ment [21,23] or whether outcome assessors were blinded

[22,23]. Two trials recruited only patients with grade IV

haemorrhoids [24,25]. One trial reported technical

difficulties during SH [26]. One study included patients

with co-morbid conditions [27].

Pain

Twenty RCTs (95%) reported less pain following SH

during the postoperative period using a VAS (Table 2).

Statistically significant heterogeneity precluded pooling

(P < 0.001; I2 = 98.5%) [21–23,28–32]. Pain lessened

over the 3 weeks postoperatively after both procedures

(Fig. 2). All 10 trials reporting VAS scores 10–15 days

postoperatively reported less pain following SH [14,25–

28,30,31,33–35].

The Bayesian meta-regression model predicted a

difference between SH and CH of 2.42 (standard error

0.31) on the VAS for pain the day after surgery. This

declined to 0.13 (SE 0.04) by day 21 (Fig. 2).

In the longer term complaints of pain were infrequent.

There was no significant difference between SH and CH

at 16 ⁄ 18 months (OR 1.03; 95% CI: 0.37, 2.88;

P = 0.95; three RCTs), or 46 months ⁄ 5 years (OR

1.84; 95% CI: 0.51, 5.52; P = 0.28; two RCTs).

Bleeding

There was no significant difference in the incidence of

bleeding during the postoperative period between SH

and CH (16 RCTs; Fig. 3) [12,22,25–28,30,32,35–42].

Significant heterogeneity was observed between studies

(P = 0.003; I2 = 57.8). The removal of the trial by Ren

[32], who reported a high incidence of bleeding after SH,

possibly due to the inclusion of patients requiring

haemostatic sutures perioperatively not counted in data

extracted from other trials, resulted in the significant

heterogeneity being eliminated (P = 0.24; I2 = 19.2%).

There remained no significant increase in bleeding

associated with SH (OR 0.86; 95% CI: 0.46, 1.61;

P = 0.63) [32]. There was no significant difference in the

incidence of postoperative haemorrhage between SH and

CH (OR 1.06; 95% CI: 0.55, 2.03; P = 0.87; 22 RCTs)

[12,14,21–26,28,30,31,33,34,36–44]. There was no

significant difference in bleeding at 12 months (OR

2.09; 95% CI: 0.91, 4.83; P = 0.08; six RCTs)

[21,24,26,33,35,44], or beyond (16 ⁄ 18 months: OR

1.84; 95% CI: 0.62, 5.50; P = 0.28 [38,43];

46 months ⁄ 5 years: OR 1.00; 95% CI: 0.33, 3.01;

P = 1.00) [34,38], between SH and CH.

Table 1 (Continued)

Study trial dates

n

SH ⁄ CH

Age

(as reported)

Male

(%)

Grades included

(as reported)

Staple gun

Comparator CH

Schmidt (2002) [40]

1998–2000

72 ⁄ 80 Range: 24–91 62% III: 123

IV: 29

Unspecified

stapling gun

Parks ⁄ Fransler-

Arnold

Senagore (2004) [35]

2001–2002

77 ⁄ 79 Mean: 49.5 (23–78) 69% III: 156 PPH 01

Ferguson

Shalaby (2001) [23]

1997–1998

100 ⁄ 100 Mean: 46.6 (SD: 13.1) 62% II: 23

III: 62

IV: 77

Prolapse: 37

PPH 01

M&M

Thaha (2003) [29]

Trial dates: NR

Thaha (2004) [49]

Trial dates: NR

48 ⁄ 42

91 ⁄ 91

Median: 50 (24–81) 58%

57%

NR Unspecified

stapling gun

Ferguson

Van de Stadt (2005) [34]

2000–2001

20 ⁄ 20 Mean: 48 (19–78) 73% II + III PPH 01

M&M

Wilson (2002) [12]

Trial dates: NR

32 ⁄ 30 Range: 40–67 NR III: 62 PPH 01

M&M

SH, stapled haemorrhoidopexy; CH, conventional excisional haemorrhoidectomy; NR, not reported; M&M, Milligan-Morgan.

Stapled haemorrhoidopexy J. Burch et al.

236 Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244

Prolapse

There was a significantly higher incidence of residual

prolapse after SH (OR 5.18; 95% CI: 1.73, 15.50;

P = 0.003; nine RCTs) [14,22–24,26,27,31,37,43].

When the trial that reported technical difficulties during

SH [26] was removed, the OR decreased to 3.38 (95%

CI: 1.00, 11.47; P = 0.05).

At 12 months, there was no significant difference

in the rate of prolapse between SH and CH (OR

3.20; 95% CI: 0.71, 14.45; P = 0.13; seven RCTs)

[21,23,24,26,35,39,44]. When data for 12 months and

beyond were pooled, prolapse was significantly more

frequent after SH (OR 4.34; 95% CI: 1.67, 11.28;

P = 0.003; 13 RCTs) [21,23,24,26,28,34,35,38,39,43–

46]. When the trials recruiting only patients with grade

Table 2 Visual analogue pain scores up to 7 days postoperatively.

Study

Number

randomized

Time point

SH mean

(SD)

CH mean

(SD)

Mean difference

(95% CI)SH CH

Ascanelli (2005) [33] 50 50 12 h 2 (NR) 7 (NR) )5

Correa-Rovelo (2002) [30] 42 42 24 h 2.8 (1.4) 5.5 (1.4) )2.70 ()3.30, )2.10)

Pavlidis (2002) [21] 40 40 24 h 0.7 (0.2) 2.4 (0.5) )1.70 ()1.87, )1.53)

Shalaby (2001) [23] 100 100 24 h 2.5 (1.3) 7.6 (0.7) )5.10 ()5.39, )4.81)

Lau (2004) [22] 13 11 Mean 2 days 3.5 (2.5) 2.6 (1.5) 0.90 ()0.72, 2.52)

Ho (2000) [28] 57 62 In hospital 4.5 (3.0) 5 (3.1) )0.50 ()1.61, 0.61)

Bikhchandani (2005) [31] 42 42 3 days 1.52 (1.43) 4.5 (2.11) )2.98 ()3.75, )2.21)

Hetzer (2002) [39] 20 20 3 days 0.8 (NR) 5.4 (NR) )4.6

Kraemer (2005) [27] 25 25 3 days 4.2 (NR) 3.7 (NR) 0.5

Krska (2003) [37] 25 25 3 days 4 (NR) 7.4 (NR) )3.4

Van de Stadt (2005) [34] 20 20 3 days 2.6 (NR) 4.7 (NR) )2.1

Boccasanta (2001) [25] 40 40 3 days 4 (NR) 6.5 (NR) )2.5

Senagore (2004) [35] 77 79 3 days 5 (NR) 6.25 (NR) )1.25

Thaha (2003) [29] 48 42 Mean 7 days 1.9 (1.58) 3.1 (1.97) )1.20 ()1.94, )0.46)

Schmidt (2002) [40] 72 80 Mean 7 days 1.83 (NR) 3.74 (NR) )1.91

Ren (2002) [32] 45 45 Unclear 2.2 (0.4) 6.4 (2.1) )4.20 ()4.82, )3.58)

Basdanis (2005) [36] 50 45 24 h 3 (1–6) 6 (3–7)

Palimento (2003) [38] 37 37 24 h 3 (1–6) 5 (3–7)

Kairaluoma (2003) [26] 30 30 3 days 3.36 (NR) 5.88 (NR)

Cheetham (2003) [14] 15 16 3 days 2.7 (NR) 7 (NR)

Chung (2005) [41] 43 45 Mean 7 days 1.5 (0.7–6) 3.5 (1.9–6)

CI, confidence interval; SD, standard deviation; SH, stapled haemorrhoidopexy; CH, conventional excisional haemorrhoidectomy; NR,

not reported.

0

1

2

3

4

5

6

7

8

9

10

0 5 10 15 20 25Number of days postoperatively

Mea

n V

AS

Sco

re

SHCHPredicted SHPredicted CH *

Figure 2 Mean visual analogue scale pain

scores reported in the included random-

ized controlled trials over the 21 day

postoperative period. *Assuming thetreatment effect declined linearly over

time.

J. Burch et al. Stapled haemorrhoidopexy

Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244 237

IV haemorrhoids [24], or reported experienced technical

difficulties during SH [26], were removed, the OR

decreased to 3.11 but was still significant (95% CI: 1.14,

8.49; P = 0.03). When high losses to follow-up were

accounted for (25% [35], 12.5% [23], 13.5% [46], 49.5%

[28,45]) the worst case scenario (all drop outs had

prolapse) resulted in no significant difference in the rate

of prolapse between SH and CH (OR 2.06; 95% CI:

0.82, 5.15; P = 0.12). This analysis was subject to

significant heterogeneity (P < 0.0002; I2 = 73.4%).

Complications

There was no significant difference in the incidence of

anastomotic stricture or anal stenosis (postoperative OR

1.15; 95% CI: 0.47, 2.79; P = 0.76 nine RCTs [26–

28,30,32,34,35,37,44]; 12 months OR 0.32; 95% CI:

0.07, 1.42; P = 0.14; four RCTs) [21,23,33,39],

or incontinence (postoperatively OR 0.73; 95%

CI: 0.35, 1.51; P = 0.39; 12 RCTs [21,22,26–

28,30,32,35,37,39–41]; 12 months OR 0.75; 95% CI:

0.26, 2.15; P = 0.59; seven RCTs [21,23,24,26,

33,35,39]; no incidents of incontinence reported beyond

12 months) [34,38,43,47].

Faecal urgency was reported after CH [35] in the

postoperative period [35,37,41]. There was no difference

at 12 months and beyond (OR 1.04; 95% CI: 0.36, 3.03;

P = 0.94; five RCTs) [21,24,33,34,43].

There was no significant difference in the incidence of

postoperative urinary retention between SH and CH (19

RCTs) [12,14,22,23,25,27,28,30,31,34–43]. This re-

mained nonsignificant when the study reporting an

incidence of 31% after SH was removed (OR: 0.76;

95% CI: 0.53, 1.09; P = 0.14) [12].

Pooled OR, where calculable, showed no significant

differences between SH and CH for anal fissure

[14,23,34,35,48], anal fistula [35,37,43], haemor-

rhoidal thrombosis [21,23–25,28,30,34,37,39,41,43],

pelvic ⁄ perianal sepsis [28,31,36,37], or rectovaginal

fistula [36] in the postoperative period. In the longer

term there were no reports of anal fistula [39],

pelvic ⁄ perianal sepsis [21,36] or rectovaginal fistula

Review:Comparison:Outcome:

Stapled haemorrhoidopexy

07 All bleeding < 4 days

Ho 2000Boccasanta 2001

Hetzer 2002Ren 2002Schmidt 2002Wilson 2002Kairaluoma 2003Krska 2003Palimento 2003Lau 2004Senagore 2004Basdanis 2005Chung 2005Kraemer 2005

Total (95% CI) 634

0/251/43

10/507/770/13

0/252/37

2/302/32

2/2028/45

2/400/261/42

2/57 0/52 5.18 5.63 [ 0.26, 119.82 ][ 0.10, 4.11 ][ 0.01, 3.08 ][ 0.12, 77.59 ][ 0.25, 123.08 ][ 8.57, 2561.67 ][ 0.13, 2.23 ][ 0.23, 108.53 ][ 0.25, 116.31 ][ 0.01, 8.25 ][ 0.18, 23.72 ]

[ 0.53, 6.68 ]

[ 0.04, 5.86 ][ 0.01, 8.25 ]

[ 0.55, 3.26 ]

[ 0.12, 0.71 ]

0.650.143.075.54

148.200.545.005.350.322.06

1.88Not estimable

0.290.510.32

1.34

8.525.114.845.095.64

10.015.145.144.806.66

10.5611.826.684.80

100.00

3/402/20

0/200/42

0/45

0/300/301/251/370/114/79

21/452/451/25

6/803/72

636Total events: 62 (Treatment), 41 (Control)Test for heterogeneity: χ2 = 33.20, df = 14 (P = 0.003), I 2 = 57.8%Test for overall effect: Z = 0.65 (P = 0.52)

0.01 0.1 1 10 100 1000Favours treament Favours control

0.001

Correa-Rovelo 2002Docherty 2001

%Weight

95% CIOR (random) OR (random)

95% CIn/N n/NControlTreatment

or sub-categoryStudy

01 Peri Post

Figure 3 Number of patients with bleeding in the immediate postoperative period.

Stapled haemorrhoidopexy J. Burch et al.

238 Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244

[21,36] after either procedure. A single incidence of

haemorrhoidal thrombosis was reported after SH at

46 months [21,34].

Wound healing

Significantly fewer patients had unhealed wounds up to

8 weeks after SH (OR 0.08; 95% CI: 0.03, 0.19;

P < 0.001; nine RCTs) [14,28,30,32,34–36,39,44].

Where reported, at 12 weeks all SH wounds had healed

[14,28,39], but 6% [14] and 20% [39] of patients had

unhealed CH wounds.

Operating time

Mean operating time ranged from 9 to 35.4 min for SH,

and 11.5–53 min for CH (19 RCTs) [21–28,30–

35,37,40,41,43,44,46]. Seventeen RCTs (89%) reported

a shorter operating time for SH [22,28]. Significant

heterogeneity precluded pooling (P < 0.001; I2 = 98.7%)

[21–23,25,26,28,30–32,41,44]. Different methods for

measuring operating time (onset of anaesthesia, time in

operating theatre, incision to application of a dressing)

may explain the observed heterogeneity.

Duration of hospital stay

The mean length of hospital stay ranged from 0.75 to

5.8 days for SH and 0.92–11.2 days for CH (16 RCTs)

[21–23,25,27,28,31–34,36,37,39,40,44,46]. Fourteen

RCTs (88%) reported a shorter hospital stay after SH

[28,39]. Significant heterogeneity precluded pooling,

even when RCTs with outlying results were excluded

from the analysis (P < 0.001; I2 = 94.0%) [32,44].

Time to work ⁄ normal activity

The mean number of days to normal activity ranged from

6.1 to 23.1 after SH and 9.8–53.9 after CH (15 RCTs)

[23,25,28,30–32,36,37,39–41,43,44,46,49]. Fourteen

RCTs (93%) reported a shorter convalescence after SH,

10 significantly so [23,25,28,30–32,36,39,41,44,46].

Significant heterogeneity precluded pooling, even when

the study reporting an unusually long convalescence after

CH [23] was removed from the analysis (P < 0.001;

I2 = 93.2%). Different definitions of return to normal

activity (return to work, period of disability) may explain

the observed heterogeneity.

Reinterventions

There was no significant difference in the total number of

reinterventions required at 12 months and beyond

between SH and CH (OR 1.74; 95% CI: 0.71, 4.24;

P = 0.23; 12 RCTs); significant heterogeneity was

observed (P = 0.08; I2 = 41.0%) [21,23,24,26,28,33–

35,39,42,43,45,46]. When one RCT recruiting only

patients with IV degree haemorrhoids [24] or technical

difficulties during SH [26], were removed, significant

heterogeneity was eliminated (P = 0.36, I2 = 9.4%). The

result remained non-significant (OR: 0.99; 95% CI: 0.47,

2.10; P = 0.98).

Reinterventions for prolapse were significantly more

frequent in the longer term after SH (OR 5.78; 95% CI:

2.0, 23.0, P = 0.002) [21,24,26,34,39,43]. Removing

the same two trials as above had little effect on the result

[24,26]. Data regarding reinterventions for bleeding,

pain and complications were sparse and inconclusive.

Patient satisfaction

Of four trials [27,28,30,31], one reported greater patient

satisfaction after SH during the postoperative period

[31]. Of eight trials [21,26,28,34,38,43–45,47], one

reported greater patient satisfaction after CH in the

longer term [34].

Cost of the procedures

Table 3 shows the mean use and cost of hospital

resources during the initial admission to be £923 after

CH and £914 after SH [mean difference £9 (SE: £10)].

The cost of the stapling gun was offset by the lower

average length of stay and theatre time for SH.

Table 3 Mean use and cost of hospital

resources during the initial admission. Unit

cost (£)

Resource use Costs (£)

CH SH CH SH

Theatre time (minutes) 8.27 29.2 15.5 242 128

Length of stay (days) 256 2.7 1.4 681 366

Device 420 0 1 0 420

Total procedure cost 923 914

SH, stapled haemorrhoidopexy; CH, conventional excisional haemorrhoidectomy.

J. Burch et al. Stapled haemorrhoidopexy

Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244 239

Discussion

This comprehensive and rigorous systematic review was

conducted with clear predefined inclusion criteria and

subgroups of interest, using extensive literature search-

ing, regular clinical advice, and established methods to

reduce error and bias during the review process. Previous

reviews included studies evaluating staplers not designed

for SH [3,20,50–54] recruited patients with thrombosed

haemorrhoids [3,20,50,53], restricted the comparator to

Milligan–Morgan and ⁄ or Ferguson [20,52,53], excluded

non-English language papers [50,54] and included a

smaller body of evidence (the Cochrane review by

Jayaraman (2006) included 12 RCTs) [51].

This review showed SH to be less painful than CH in

the postoperative period, with no associated increase in

bleeding but a higher rate of residual prolapse. SH was

associated with shorter operating times, hospital stay, and

convalescence, and fewer unhealed wounds. There was no

difference in the rate or type of complications between

SH and CH.

The incidence of recurrent prolapse following SH

remains contentious amongst surgeons, potentially pre-

venting its full acceptance as a credible alternative to

CH. Early prolapse observed in this review may be a

result of residual skin tags being misinterpreted as

persisting prolapse [55]. Late prolapse, however, is more

likely to be due to failure of the stapling technique to

remove an adequate volume of prolapsing tissue. The

PPH-01 stapler accommodates a set volume of tissue

within its housing. If this is exceeded inadequate

haemorrhoidal excision is likely to occur, leaving the

patient susceptible to recurrent symptoms. This is

supported by the decrease in the OR for recurrent

prolapse when grade IV haemorrhoids are removed

from our analysis. Where large volume haemorrhoids are

encountered, a double stapling procedure may result in

a more complete haemorrhoidal reduction with a

corresponding reduction in the incidence of recurrent

prolapse [55].

On average the difference in hospital costs between

the procedures was £9. The additional cost of the

staple gun was largely offset by savings in operating

time and hospital stay. The difference in cost is thus

marginal and unlikely to prohibit the clinical use of SH.

Further modelling is required to evaluate the costs and

QALYs, and the cost-effectiveness of the strategies over

a longer time horizon. Given the decreased wound

complications observed following SH it is likely that

the community costs will be favourable to the stapling

technique.

Much of the variability between studies seemed to be

related to the grade of haemorrhoids or the apparent

experience of the surgeons. A study that recruited only

patients with grade IV haemorrhoids seemed responsible

for the heterogeneity in the analysis of the rate of

reintervention [24]. A trial reporting technical difficulties

during SH seemed responsible for the heterogeneity in

the analysis of residual prolapse and the requirement for

reintervention [26].

Limitations

In this review of the available data, all included studies

had methodological flaws. There were no large, high

quality RCTs conducted in a representative population.

Several trials were small and possibly underpowered.

Data for long-term outcomes were limited and often

subject to a high incidence of loss to follow-up. A

prospective register of 810 patients who underwent SH

in 2005 has been compiled under the auspices of the

Association of Coloproctology of Great Britain and

Ireland with post operative data collected at 6 weeks of

follow up (Mr M. Lamparelli, personal communication,

June 2007). Continued follow-up of the patients

registered may provide information regarding the

long-term effectiveness and reintervention rates follow-

ing SH.

The main limitation of any economic study is the

lack of directly observed utility data in the early recovery

period, making difficult the expression of differences

between the procedures. The indirect methods used to

estimate utilities require key assumptions to be made.

The results are sensitive to modelling assumptions,

particularly the valuation of utility in the early postop-

erative period. Therefore, when discussing the intro-

duction of SH, NHS managers need to assess the

potential for shortening stays, by reducing the length of

inpatient admissions or increasing the proportion of day

cases.

Future research

The lack of long-term data means that the magnitude of

the increased rate of prolapse and reintervention associ-

ated with SH is unclear. Ideally an adequately powered

good quality RCT comparing SH and CH to include

patients with stages II, III and IV degree haemorrhoids

with prolapse being established by physical examination

and with at least 5 years follow-up is recommended.

Follow-up of the patients registered on the recently

compiled prospective register should continue. Further

research requirements include follow up to at least

6 months postoperatively, the ability of the technique

to reduce hospital stay and a systematic review of all

available treatments for haemorrhoids.

Stapled haemorrhoidopexy J. Burch et al.

240 Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244

Conclusions

Stapled haemorrhoidopexy resulted in less pain in the

postoperative period, but a higher rate of prolapse, and

the need for reintervention for prolapse in the longer

term. There was no clear difference in the rate or type of

complications between the two techniques. On average

the difference in hospital costs was £9; additional cost of

the staple gun was offset by savings in operating time and

hospital stay. Given the currently available evidence, the

decision to conduct SH or CH should primarily be based

upon the priorities and preferences of the patient (short-

term reduction in pain and convalescence, or longer term

reduced risk of recurrence) and surgeon. An adequately

powered RCT comparing SH with CH, in patients with

II, III and IV degree haemorrhoids, with at least 5 years

follow-up is recommended, dependent upon the results

of the prospective register.

Conflict of interest

None.

Acknowledgements

We would like to thank Professor Mark Sculpher,

Professor John Monson, Dr Ken Stein, Su Golder and

Professor Mike Drummond for their assistance during

the review. We would like to thank those authors who

responded to our requests for further information. We

retrieved several foreign language papers and would also

like to thank our translators.

References

1 Shanmugam V, Thaha MA, Rabindranath KS, Campbell KL,

Steele RJC, Loudon MA. Rubber band ligation versus

excisional haemorrhoidectomy for haemorrhoids. Cochrane

Database Syst Rev 2005; Issue 1. Art. No. CD005034. DOI:

10.1002/14651858.CD005034.pub2.

2 Lacerda-Filho A, Da Silva RG. Stapled hemorrhoidectomy:

present status. Arq Gastroenterol 2005; 42: 191–4.

3 Nisar PJ, Scholefield JH. Managing haemorrhoids. BMJ

2003; 327: 847–51.

4 Johannsson HO. Haemorrhoids: aspects of symptoms and

results after surgery. Acta Univ Ups 2005; 86: 90.

5 Madoff RD, Fleshman JW. American Gastroenterological

Association technical review on the diagnosis and treatment

of hemorrhoids. Gastroenterology 2004; 126: 1463–73.

6 Goligher JC. (1984) Surgery of the Anus, Rectum and Colon,

5th edn. Bailliere Tindall, London.

7 NHS Information Centre. (2004 ⁄ 05) HESonline (Hospital

Episode Statistics). [accessed 31 ⁄ 07 ⁄ 06]; http://www.

hesonline.nhs.uk.

8 Longo A. (1998) Treatment of hemorrhoids disease by

reduction of mucosa and haemorrhoidal prolapse with a

circular suturing device: a new procedure. 6th World Congress

of Endoscopic Surgery (IFSES). Rome 1998 June 3–6.

9 Eu KW, Lai JH. Stapled haemorrhoidectomy or Longo�sprocedure? Two totally different concepts. Singapore Med J

2005; 46: 566–7.

10 Monson JRT, Mortenson NJ, Hartley J. (2002) Procedures

from Prolapsing Haemorrhoids (PPH) or Stapled Anopexy.

Consensus Document for Association of Coloproctology of Great

Britain and Ireland (ACPGBI): ACPGBI, London.

11 Lehur PA, Gravie JF, Meurette G. Circular stapled anopexy

for haemorrhoidal disease: results. Colorectal Dis 2001; 3:

374–9.

12 Wilson MS, Pope V, Doran HE, Fearn SJ, Brough WA.

Objective comparison of stapled anopexy and open hemor-

rhoidectomy: a randomized, controlled trial. Dis Colon

Rectum 2002; 45: 1437–44.

13 George BD, Shetty D, Lindsey I, Mortensen NJM, Warren

BF. Histopathology of stapled haemorrhoidectomy speci-

mens: a cautionary note. Colorectal Dis 2002; 4: 473–6.

14 Cheetham MJ, Cohen CR, Kamm MA, Phillips RK. A

randomized, controlled trial of diathermy hemorrhoidecto-

my vs. stapled hemorrhoidectomy in an intended day-care

setting with longer-term follow-up. Dis Colon Rectum 2003;

46: 491–7.

15 Molloy RG, Kingsmore D. Life threatening pelvic sepsis after

stapled haemorrhoidectomy. Lancet 2000; 355: 810.

16 Burch J, Epstein D, Baba-Akbari A, Weatherly H, Fox D,

Golder S, Jayne D, Drummond M, Woolacott N. Stapled

haemorrhoidectomy (haemorrhoidopexy) for the treatment

of haemorrhoids: a systematic review and economic evalua-

tion. Health Technol Assess 2008; 12: 8.

17 Centre for Reviews & Dissemination. CRD Report No. 4.

(2001) Undertaking Systematic Reviews of Research on

Effectiveness: CRD Guidelines for Those Carrying out or

Commissioning Reviews, 2nd edn. Centre for Reviews and

Dissemination, York.

18 Sutton AJ, Abrams KR, Jones D, Sheldon T, Song F. (2000)

Methods for Meta-Analysis in Medical Research. Wiley,

Chichester.

19 Department of Health. (2005) NHS Reference Costs 2005.

[cited 21 May 2007]; http://www.dh.gov.uk/en/

Publicationsandstatistics/Publications/PublicationsPolicy

And Guidance/DH_4133221.

20 Ethicon Endo-Surgery Inc. Stapled haemorrhoidopexy for

the treatment of haemorrhoids (submission to the National

Institute for Health & Clinical Excellence). Ethicon Endo-

Surgery Inc 2006.

21 Pavlidis T, Papaziogas B, Souparis A, Patsas A, Koutelidakis I,

Papaziogas T. Modern stapled Longo procedure vs. conven-

tional Milligan-Morgan hemorrhoidectomy: a randomized

controlled trial. Int J Colorectal Dis 2002; 17: 50–53.

22 Lau PYY, Meng WCS, Yip AWC. Stapled haemorrhoidecto-

my in Chinese patients: a prospective randomised control

study. Hong Kong Med J 2004; 10: 373–7.

J. Burch et al. Stapled haemorrhoidopexy

Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244 241

23 Shalaby R, Desoky A. Randomized clinical trial of stapled

versus Milligan-Morgan haemorrhoidectomy. Br J Surg

2001; 88: 1049–53.

24 Ortiz H, Marzo J, Armendariz P, De Miguel M. Stapled

hemorrhoidopexy vs. diathermy excision for fourth-degree

hemorrhoids: a randomized, clinical trial and review of the

literature. Dis Colon Rectum 2005; 48: 809–15.

25 Boccasanta P, Capretti PG, Venturi M, Cioffi U, De Simone

M, Salamina G et al. Randomised controlled trial between

stapled circumferential mucosectomy and conventional cir-

cular hemorrhoidectomy in advanced hemorrhoids with

external mucosal prolapse. Am J Surg 2001; 182: 64–68.

26 Kairaluoma M, Nuorva K, Kellokumpu I. Day-case stapled

(circular) vs. diathermy hemorrhoidectomy: a randomized,

controlled trial evaluating surgical and functional outcome.

Dis Colon Rectum 2003; 46: 93–99.

27 Kraemer M, Parulava T, Roblick M, Duschka L, Muller-

Lobeck H. Prospective, randomized study: proximate PPH

stapler vs. LigaSure for hemorrhoidal surgery. Dis Colon

Rectum 2005; 48: 1517–22.

28 Ho YH, Cheong WK, Tsang C, Ho J, Eu KW, Tang CL

et al. Stapled hemorrhoidectomy - cost and effectiveness.

Randomized, controlled trial including incontinence scoring,

anorectal manometry, and endoanal ultrasound assessments

at up to three months. Dis Colon Rectum 2000; 43: 1666–

75.

29 Thaha MA, Kazmi SA, Binnie NR, Hendry WS, Campbell

KL, Steele RJC. Pain following haemorrhoid surgery:

preliminary results of a randomised controlled trial compar-

ing circular stapled anopexy and closed haemorrhoidectomy

performed in the prone position. Br J Surg 2003; 90: 68.

30 Correa-Rovelo JM, Tellez O, Obregon L, Miranda-Gomez

A, Moran S. Stapled rectal mucosectomy vs. closed hemor-

rhoidectomy: a randomized, clinical trial. Dis Colon Rectum

2002; 45: 1367–74.

31 Bikhchandani J, Agarwal PN, Kant R, Malik VK. Random-

ized controlled trial to compare the early and mid-term

results of stapled versus open hemorrhoidectomy. Am J Surg

2005; 189: 56–60.

32 Ren DL, Luo ZB, Zhang SF, Fan XH, Zao JL, Feng Y.

Procedure for proplapse and hemorrhoids (PPH) versus

Milligan-Morgan hemorrhoidectomy: a randomised con-

trolled study. Chin J Curr Adv Gen Surg 2002; 17: 714–5.

33 Ascanelli S, Gregorio C, Tonini G, Baccarini M, Azzena G.

Long stapled haemorrhoidectomy versus Milligan-Morgan

procedure: short- and long-term results of a randomised,

controlled, prospective trial. Chir Ital 2005; 57: 439–47.

34 Van de Stadt J, D�Hoore A, Duinslaeger M, Chasse E,

Penninckx F. Long-term results after excision haemorrhoi-

dectomy versus stapled haemorrhoidopexy for prolapsing

haemorrhoids; a Belgian prospective randomized trial. Acta

Chir Belg 2005; 105: 44–52.

35 Senagore AJ, Singer M, Abcarian H, Fleshman J, Corman M,

Wexner S et al. A prospective, randomized, controlled

multicenter trial comparing stapled hemorrhoidopexy and

Ferguson hemorrhoidectomy: perioperative and one-year

results. Dis Colon Rectum 2004; 47: 1824–36.

36 Basdanis G, Papadopoulos VN, Michalopoulos A, Apostol-

idis S, Harlaftis N. Randomized clinical trial of stapled

hemorrhoidectomy vs open with Ligasure for prolapsed piles.

Surg Endosc 2005; 19: 235–9.

37 Krska Z, Kvasnieka J, Faltyn J, Schmidt D, Svab J, Korma-

nova K et al. Surgical treatment of haemorrhoids according

to Longo and Milligan Morgan: an evaluation of postoper-

ative tissue response. Colorectal Dis 2003; 5: 573–6.

38 Palimento D, Picchio M, Attanasio U, Lombardi A, Bambini

C, Renda A. Stapled and open hemorrhoidectomy: Ran-

domized controlled trial of early results. World J Surg 2003;

27: 203–7.

39 Hetzer FH, Demartines N, Handschin AE, Clavien PA.

Stapled vs excision hemorrhoidectomy: long-term results of a

prospective randomized trial. Arch Surg 2002; 137: 337–40.

40 Schmidt MP, Fischbein J, Shatavi H. Stapler hemorrhoidec-

tomy versus conventional procedures – a clinical study.

Zentralbl Chir 2002; 127: 15–18.

41 Chung CC, Cheung HY, Chan ES, Kwok SY, Li MK.

Stapled hemorrhoidopexy vs. harmonic scalpel hemorrhoi-

dectomy: a randomized trial. Dis Colon Rectum 2005; 48:

1213–9.

42 Docherty JG, Sunderland GT, Anderson JH. Prospective

randomised trial of closed haemorrhoidectomy and stapled

prolapsectomy for haemorrhoids: one year follow up

[abstract]. Colorectal Dis 2001; 3(Suppl. 1): Poster 124.

43 Ortiz H, Marzo J, Armendariz P. Randomized clinical trial of

stapled haemorrhoidopexy versus conventional diathermy

haemorrhoidectomy. Br J Surg 2002; 89: 1376–81.

44 Hasse C, Sitter H, Brune M, Wollenteit I, Lorenz W,

Rothmund M. Haemorrhoidectomy: conventional excision

versus resection with the circular stapler. Prospective,

randomized study. Dtsch Med Wochenschr 2004; 129:

1611–7.

45 Ooi BS, Ho YH, Tang CL, Eu KW, Seow-Choen F. Results

of stapling and conventional hemorrhoidectomy. Tech Col-

oproctol 2002; 6: 59–60.

46 Gravie JF, Lehur PA, Huten N, Papillon M, Fantoli M,

Descottes B et al. Stapled hemorrhoidopexy versus Milligan-

Morgan hemorrhoidectomy: a prospective, randomized,

multicenter trial with 2-year postoperative follow up. Ann

Surg 2005; 242: 29–35.

47 Picchio M, Palimento D, Attanasio U, Renda A. Stapled vs

open haemorrhoidectomy: long-term outcome of a random-

ized controlled trial. Int J Colorectal Dis 2006; 21: 668–9.

48 Accarpio G, Ballari F, Puglisi R, Menoni S, Ravera G,

Accarpio FT et al. Outpatient treatment of hemorrhoids with

a combined technique: results in 7850 cases. Tech Coloproctol

2002; 6: 195–6.

49 Thaha MA, Kazmi SA, Binnie NR, Hendry WS, Staine HJ,

Campbell KL et al. Duration of pain and its influence on

return to work following haemorrhoid surgery: results of

multi-centre randomized controlled trial comparing circular

stapled anopexy and Ferguson closed haemorrhoidectomy.

Br J Surg 2004; 91: 2.

50 Sutherland LM, Burchard AK, Matsuda K, Sweeney JL,

Bokey EL, Childs PA, et al. (2002) A Systematic Review of

Stapled haemorrhoidopexy J. Burch et al.

242 Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland No claim to original US government works. Colorectal Disease, 11, 233–244

Stapled Hemorrhoidectomy. ASERNIP-S, Adelaide, Report

No.: 0909844453.

51 Jayaraman S, Colquhoun PHD, Malthaner RA. Stapled

versus conventional surgery for hemorrhoids. Cochrane

Database of Systematic Reviews 2006; issue 4. Art. No.:

CD005393. DOI: 10.1002/14651858.CD005393.pub2.

52 Nisar PJ, Acheson AG, Neal KR, Scholefield JH. Stapled

hemorrhoidopexy compared with conventional hemorrhoi-

dectomy: systematic review of randomized, controlled trials.

Dis Colon Rectum 2004; 47: 1837–45.

53 Lan P, Wu X, Zhou X, Wang J, Zhang L. The safety and

efficacy of stapled hemorrhoidectomy in the treatment of

hemorrhoids: a systematic review and meta-analysis of ten

randomized control trials. Int J Colorectal Dis 2006; 21:

172–8.

54 Tjandra JJ, Chan MKY. Systematic review on the procedure

for prolapse and hemorrhoids (stapled hemorrhoidopexy).

Dis Colon Rectum 2007; 50: 878–92.

55 Boccasanta P, Venturi M, Roviaro G. Stapled transanal rectal

resection versus stapled anopexy in the cure of hemorrhoids

associated with rectal prolapse. A randomized controlled

trial. Int J Colorectal Dis 2007; 22: 245–51.

Commentary doi:10.1111/j.1463-1318.2008.01760.x

Conventional haemorrhoidectomy is an effective tech-

nique for dealing with haemorrhoidal prolapse, bleeding

and perianal skin tags. Patient selection and operative

technique are important and the outcome is usually

satisfactory. However, open haemorrhoidectomy is

uncomfortable, even painful and on average is associated

with a 2-week period of convalescence before return to

work. Efforts to minimize discomfort and accelerate

healing, through the use of perioperative antibiotics,

primary closure, submucosal resection and various energy

sources to facilitate haemorrhoid excision have largely

proven unsuccessful in providing a better outcome when

compared with the classical Milligan and Morgan

operation.

In the past decade, Dr Antonio Longo�s stapling

technique, introduced in 1998 [1], focused on mucosal

prolapse as a pivotal component in the evolution of

symptomatic haemorrhoids. While the concept itself

was not new [2,3], adaptation of available circular

stapling instruments to effectively reposition haemor-

rhoidal tissue above the dentate line has been a major

advance. The operation PPH (procedure for prolapse

and haemorrhoids) or stapled haemorrhoidopexy [4], as

it became known, has been shown repeatedly to be less

painful than conventional haemorrhoidectomy. As a

consequence, postoperative hospital stay has generally

been shorter and return to work quicker. Initial

concerns of potentially serious postoperative complica-

tions have proven unfounded. Subsequent large case

series [5] and systematic reviews [6,7] have confirmed

stapled haemorrhoidopexy as a safe and effective

procedure in the treatment of prolapsing internal

haemorrhoids albeit with a somewhat higher recurrence

rate [6].

One might think therefore that a role for stapled

haemorrhoidopexy has been established and that the

procedure should become a widely available alternative to

open haemorrhoidectomy for patients judged suitable by

appropriately trained surgeons. Not so; the new opera-

tion comes with a cost. Stapling is much more expensive

than sutures! The current systematic review by Burch

et al. arises from a Health Technology Assessment

exercise conducted for the UK National Health Service

to determine (yet again) whether the treatment is

effective but also, and more importantly, how cost

effective the treatment is. The approach taken is primarily

from a health economics perspective based on data

abstracted from available randomized clinical trials. Not

surprisingly, the findings are similar to other systematic

reviews. What is novel is the application of a relatively

simple costing analysis to the data available from 27 trials

and 2279 patients. This consisted of apportioning a cost

to theatre time, length of hospital stay and the additional

cost of the device in the stapled group.

The conclusion that the overall cost difference

between the procedures is £9 with additional device

costs offset by shorter hospital stay is welcome as is the

view expressed that the decision to conduct a stapled

haemorrhoidopexy or conventional haemorrhoidectomy

should primarily be based upon the priorities and preferences

of the patient and surgeon. Inherent in this statement, and

in the earlier discussion that NHS managers need to assess

the potential for shortening stays (i.e. cost containment), is

the implication that were the cost benefit analysis

somewhat different, then cost containment might take

priority and access to the technology might be restricted.

There is no doubt that in today�s more straightened

economic circumstances, both clinicians and managers

P. Ronan O’Connell Commentary

� 2009 The Authors. Journal Compilation � 2009 The Association of Coloproctology of Great Britain and Ireland. Colorectal Disease, 11, 233–244 243