International Journal of Gynecology and Obstetrics 110 (2010) 27–30

Contents lists available at ScienceDirect

International Journal of Gynecology and Obstetrics

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j g o

CLINICAL ARTICLE

Risk factors for vaginal prolapse after hysterectomy

Adolf Lukanovič a,⁎, Katarina Dražič b
a
Department of Obstetrics and Gynecology, University Medical Center Ljubljana, Ljubljana, Slovenia
b
Department of Obstetrics and Gynecology, General Hospital Jesenice, Slovenia

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To identify risk factors for pelvic organ prolapse (POP) and their influence on the occurrence of
Received 30 November 2009 vaginal prolapse after hysterectomy. Methods: Medical records from 2 groups of women who had undergone
Received in revised form 26 January 2010 hysterectomy were reviewed retrospectively. The study group was 82 women who had undergone surgery
Accepted 1 March 2010 for vaginal prolapse after hysterectomy; the control group was 124 women who had undergone
hysterectomy with no diagnosis of vaginal prolapse by the time of the study. All hysterectomy procedures
Keywords:
had been performed for benign gynecological disease, including POP. Both groups of women completed a
Hysterectomy
Pelvic organ prolapse
self-administered questionnaire to obtain additional information on the occurrence of POP. Results: The
Risk factors incidence of vaginal prolapse after hysterectomy was significantly higher in women with a higher number of
vaginal deliveries, more difficult deliveries, fewer cesareans, complications after hysterectomy, heavy
physical work, neurological disease, hysterectomy for pelvic organ prolapse, and/or a family history of pelvic
organ prolapse. Premenopausal women had vaginal prolapse corrected an average of 16 years after
hysterectomy, and postmenopausal women 7 years post hysterectomy. Conclusion: Before deciding on
hysterectomy as the approach to treat a woman with pelvic floor dysfunction, the surgeon should evaluate
these risk factors and discuss them with the patient.
© 2010 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction time elapsed since hysterectomy, from 1% at 3 years after hysterec-

tomy to 5% at 15 years after hysterectomy [5].
Genital prolapse is a common gynecological condition that occurs It is questionable whether POP is more common after vaginal than
predominantly in middle and old age. Its true incidence is unknown, after abdominal hysterectomy, but Morely and DeLancey [6] state that
and it frequently manifests itself soon after menopause. Little is known POP occurs equally often after either type of hysterectomy. Some
about the risk factors for genital prolapse and specifically about the studies in the literature suggest that the risk of POP differs between
risk factors for genital prolapse after hysterectomy, one of the most ethnic groups [7–9].
common surgical procedures performed in women. The aim of the present retrospective study was to investigate the
The International Continence Society defines posthysterectomy relationship between some known risk factors for the development
(apical) vaginal prolapse as descent of the vaginal cuff scar below a of POP and the incidence of posthysterectomy POP in a population
point that is 2 cm less than the total vaginal length above the plane of of women from central Europe. The data obtained from a self-
the hymen [1]. administered questionnaire and from medical records of women who
It is not known whether vaginal posthysterectomy prolapse is the had undergone a hysterectomy and had required surgical vaginal
consequence of pelvic floor relaxation or of a genetic predisposition or prolapse repair after the hysterectomy were compared with data from
how this problem could be postponed or avoided [2]. Hysterectomy a control group of women who by the start of the study period had not
is the second most frequently performed surgical procedure (after required a vaginal prolapse repair after hysterectomy performed for a
cesarean delivery) in women in the USA [3] and more than 30% of benign gynecological disease.
hysterectomies are performed in women younger than 60 years of
age; pelvic organ prolapse (POP) is the indication for surgery in 33% 2. Materials and methods
of women after the age of 55 years [4]. Mant et al. [5] reported that
the incidence of posthysterectomy POP that requires surgical repair is In 2004, a national multicenter retrospective matched case-control
3.6 per 1000 cases. The cumulative risk for surgery increases with the study was carried out. The study was designed to compare 2 groups of
women after hysterectomy. The study group consisted of 121 women
who had undergone surgical correction of POP after hysterectomy at a
⁎ Corresponding author. Department of Obstetrics and Gynecology, University
Medical Centre Ljubljana, Šlajmerjeva 3, SI-1000 Ljubljana, Slovenia. Tel.: + 386 1
university department of obstetrics and gynecology between January
522 62 00; fax: + 386 1 522 61 30. 1, 2000 and December 31, 2003. The control group consisted of 241
E-mail address: [email protected] (A. Lukanovič). women who had undergone a hysterectomy and had not developed

0020-7292/$ – see front matter © 2010 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.ijgo.2010.01.025
28 A. Lukanovič, K. Dražič / International Journal of Gynecology and Obstetrics 110 (2010) 27–30

a vaginal prolapse by the time of data collection. The women in the Table 1
control group were selected from medical charts and were matched Characteristics of the study population at hysterectomy.a

to the women in the study group by age, parity, menopausal status, Characteristics Study group Control group P value b

hospital at which their hysterectomy had been performed, year (n = 82) (n = 124)
of surgery, indication, and type of surgery. The operations were POP as the indication for hysterectomy 46 (56.1) 50 (40.3) 0.026
performed in 10 different gynecological departments across Slovenia. Age, y 52.2 ± 9.5 53.3 ± 10.2 0.437 c

c
All hysterectomy procedures were performed for benign gynecological Body weight, kg 68.7 ± 10.5 70.3 ± 10.4 0.283
d
Vaginal deliveries 2.3 ± 1.1 1.9 ± 1.2 0.007
pathology according to the standard surgical protocol. The surgical
No. of cesarean deliveries
technique included plication of uterosacral ligaments to prevent 0 82 (100) 113 (91.1) 0.021
postoperative complications. We obtained additional data from the 1 0 (0) 9 (7.3)
hospital charts of women who had follow-up examinations, which 2 0 (0) 2 (1.6)
included pelvic examination by the surgeon who had performed the No. of difficult deliveries
0 40 (48.8) 80 (64.5) 0.020
hysterectomy.
1 31 (37.8) 35 (28.2)
A postal questionnaire was sent to all of the women enrolled in 2 11 (13.4) 9 (7.3)
the study to obtain additional information on the period after the Postoperative complications 18 (22.0) 12 (9.7) 0.015
hysterectomy. The questionnaire required provision of demographic Type of work
Heavy 14 (17.1) 9 (7.3) 0.027
and personal data on age at hysterectomy; body weight at the time of
Not heavy 43 (52.4) 59 (47.6)
hysterectomy; increase in body weight from the time of hysterectomy Light 25 (30.5) 56 (45.2)
to the time of vaginal prolapse repair (study group) or to the time of Neurological disease 7 (8.5) 2 (1.6) 0.017
data collection (control group); number of vaginal deliveries; number Family history of pelvic organ prolapse 34 (41.5) 28 (22.6) 0.004
of cesarean deliveries; number of difficult deliveries (protracted labor, Abbreviations: POP, pelvic organ prolapse.
instrumental delivery, neonatal birth weight N4000 g, perineal tear); a
Values are given as number (percentage) or mean ± standard deviation unless
menopausal status; type of hysterectomy (vaginal or abdominal); otherwise indicated.
b
P values were calculated using the χ2 test unless otherwise indicated.
consequences of hysterectomy (pain and pressure sensation, urinary c
Calculated using the t test.
retention, urinary incontinence, fecal incontinence); symptoms of d
Calculated using the Mann-Whitney test.
vaginal prolapse after the hysterectomy (pelvic discomfort, pressure,
pain); type of work; neurological disease; and a family history of POP.
The sole inclusion criterion for the study was that hysterectomy complications (P = 0.015), heavy work (P = 0.027), vaginal deliveries,
had been performed in both groups for benign pathology (myomas, neurological disease (P = 0.017), and family history of POP (P = 0.004)
bleeding, POP, endometriosis, adenomyosis, chronic pelvic pain, in the study group. In the study group, none of the women had had a
infection, or endometrial hyperplasia). Exclusion criteria were hyster- cesarean delivery, whereas in the control group 9 (7.3%) women had
ectomy performed for malignant pathology and hysterectomy at had 1 cesarean delivery, and 2 (1.6%) women had had 2 cesarean
the time of cesarean delivery. The inclusion criterion for POP repair deliveries. Age and body weight did not differ between the groups.
was any prolapse of the vaginal cuff below the hymeneal remnants. Of the 46 (56.1%) women with POP as the indication for hysterectomy
Women were excluded from the control group if they reported on the in the study group, 25 (54.3%) women were postmenopausal. Of the 50
questionnaire that they had undergone surgery for POP in 1 of 10 (40.3%) women in the control group with POP as the indication for
gynecologic departments in Slovenia. Medical records revealed normal hysterectomy, 41 (82.0%) were postmenopausal.
pelvic status at the first 3 follow-up examinations after hysterectomy in In the study group, 40 (48.8%) women had an abdominal hysterec-
all women enrolled in the study (6 weeks, 6 months, and 1 year after tomy and 42 (51.2%) had a vaginal hysterectomy. In the control group,
hysterectomy). 57 (46.0%) women had an abdominal hysterectomy and 67 (54%) had a
Heavy work was defined as a job involving both long hours vaginal hysterectomy. There was no statistically significant difference
primarily in a standing position and lifting objects heavier than 5 kg. between the groups.
The term difficult delivery encompassed prolonged labor, perineal In the study group, the incidence of postoperative complications
tear of third degree or greater, forceps delivery, vacuum extraction, or after hysterectomy was significantly higher than in the control group
a birth weight greater than 4000 g. (Table 2).
The study was approved by the National Medical Ethics Committee. Among the study group, women who had undergone hysterecto-
The women in the study and control groups were given a detailed my because of POP had a vaginal prolapse repair an average of 9 years
explanation of the study objectives and provided written consent for after hysterectomy, whereas the women who underwent hysterecto-
participation in the study. my because of other benign pathology had vaginal prolapse repair an
Descriptive statistics (t test, χ2 test, Mann-Whitney test) and average of 18 years after hysterectomy (P b 0.001).
multivariate logistic regression models were used to investigate the With regard to postmenopausal status, women in the study group
factors that were associated with POP. Statistical analysis was who had undergone hysterectomy for POP and were postmenopausal
performed using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA). at the time of the procedure underwent vaginal prolapse repair an
Differences were considered significant when P b 0.05. average of 7 years after primary hysterectomy; whereas the women
who were still menstruating at the time of hysterectomy underwent
3. Results
Table 2
The study and control groups were homogeneous for age, number Postoperative complications.a
of deliveries, menopausal status at hysterectomy, and type of b
Complications Study group Control group P value
hysterectomy. Postal questionnaires were sent to 121 women in the (n = 82) (n = 124)
study group and 241 women in the control group, and 82 (67.8%) and
Pain and pressure sensation 40 (48.8) 0 b0.001
124 women (51.5%) responded, respectively. Urinary incontinence 31 (37.8) 13 (10.5) b0.001
The characteristics of the patients at hysterectomy are presented in Urinary retention 13 (15.9) 1 (0.8) b0.001
Table 1. Comparison between women in the study and control groups Fecal incontinence 6 (7.3) 0 0.002
demonstrated significantly higher rates of POP as the indication for a
Values are expressed as number (percentage).
hysterectomy (P = 0.026), difficult delivery (P = 0.020), postoperative b
P values were calculated using the χ2 test.
 A. Lukanovič, K. Dražič / International Journal of Gynecology and Obstetrics 110 (2010) 27–30 29

Table 3 assessment of clinical applicability of diagnostic tests. The surgeon

Multivariate analysis. Odds ratios and 95% confidence intervals adjusted for other who treats POP of any degree must be able to tailor the surgery to meet
predictive factors of pelvic organ prolapse in women who had undergone a
hysterectomy.
each patient's needs. This mandates proficiency in the performance of
a number of surgical approaches and precludes adherence to a single
Variable B P value Odds ratio (95% CI) approach. Current surgical procedures designed to correct pelvic floor
POP indication for hysterectomy 1.03 0.01 2.80 (1.30–6.01) defects, including POP, need to follow 2 basic principles: to reapproach
Heavy work 1 (Reference) the preserved structures and to use various types of mesh in case the
No heavy work –0.80 0.12 0.45 (0.16–1.24)
tissue has become dysfunctional.
Light work –1.09 0.04 0.34 (0.12–0.95)
Uncomplicated delivery 0.18 0.23 1.19 (0.90–1.59) Several procedures are recommended to prevent vaginal prolapse
Difficult delivery 0.51 0.05 1.67 (1.01–2.78) at the time of hysterectomy. One procedure is McCall culdoplasty, a
Age at hysterectomy –0.04 0.03 0.96 (0.92–1.00) recommended measure to prevent enterocele formation at the time of
Family history of POP 0.81 0.015 2.2 (1.2–4.3) vaginal hysterectomy [14–17]; another procedure involves suturing
Abbreviations: B, logistic regression coefficient; CI, confidence interval; POP, pelvic the cardinal and uterosacral ligaments to the vaginal cuff [18]; and a
organ prolapse. further procedure is sacrospinous fixation at the time of vaginal
hysterectomy when the vault descends to the introitus during closure
[19]. However, no comparative studies are available to assess the true
vaginal prolapse repair an average of 16 years after hysterectomy value of these preventive procedures.
(P b 0.001). In the present study, there was a significant difference in the time
The most important risk factors in the multivariate analysis were interval between hysterectomy and the vaginal prolapse repair ac-
POP as the indication for hysterectomy (adjusted OR 2.8; 95% CI, 1.30– cording to the women's menopausal status. Women who had a
6.01) and family history of POP (adjusted OR 2.2; 95% CI, 1.2–4.3) hysterectomy for POP and were postmenopausal had a vaginal
(Table 3). Other statistically significant factors were difficult deliver- prolapse correction done an average of 7 years later, whereas those
ies, heavy work, and younger age at hysterectomy (Table 3). who were premenopausal at the time of hysterectomy for POP had a
repair an average of 16 years after hysterectomy (P b 0.001). When
4. Discussion estimating the impact of menopausal status in women who had
undergone hysterectomy for POP, we found that 25 (54.3%) women in
The true incidence of vaginal prolapse after hysterectomy is the study group and 41 (82%) women in the control group were
unknown, mainly because of the different classifications used. Data in postmenopausal. These results could indicate the role of circulating
the literature are scarce and not uniform. Mattingly and Thompson [10] estrogen levels and their impact on the synthesis of collagen type 3,
reported that the incidence of severe vaginal vault prolapse after which is responsible for tissue elasticity.
hysterectomy was less than 0.5%, whereas Birnbaum [11] reported In line with a published study [20], we did not find any difference
the incidence of posthysterectomy vaginal vault prolapse as less than in the incidence of vaginal prolapse between women who underwent
1%. In their epidemiological study of POP, Mant et al. [5] estimated the hysterectomy by the vaginal or abdominal route.
incidence of severe vaginal vault prolapse after hysterectomy to range Pregnancy and delivery are thought to be the main cause of POP—
between 5% and 29%. Their results showed the risk of vaginal prolapse whether the prolapse occurs immediately after pregnancy or 30 years
after hysterectomy was 5.5 times higher among women whose initial later. High birth weight of the newborn and the physical trauma of
hysterectomy was for genital prolapse as opposed to other indications. labor and birth both stress and strain the pelvic muscles, and liga-
Olsen et al. [4] found that the incidence of posthysterectomy surgery ments might become damaged and never fully regain their strength
for vaginal prolapse was higher in women who had undergone and elasticity. Mant et al. [5] found that after 4 vaginal deliveries
hysterectomy owing to POP (158 per 10 000 women) than in women the risk of developing POP is 11 times higher. A positive correlation
who had undergone hysterectomy for other benign pathologies between multiparity, birth weight (N4500 g), and the incidence of
(29 per 10 000 women). These results are also in agreement with the POP was also found in the study by Samuelsson [21].
retrospective follow-up study by Marchioni et al. [12] of 448 women In the present study, we also found that women in the study
who underwent hysterectomy. The study found that vaginal vault group had a significantly higher number of vaginal deliveries, fewer
prolapse followed 11.6% of hysterectomies performed for vaginal cesarean deliveries, and a significantly higher number of difficult
prolapse and 1.8% of hysterectomies performed for other indications. deliveries. It is possible that these factors influence the incidence of
Swift and Theofrastous [13] also found a significant correlation between posthysterectomy vaginal prolapse.
vaginal prolapse and a previous hysterectomy performed for POP. We did not investigate the influence of obesity or chronic cough on
The results of the present study also confirm the correlation the posthysterectomy incidence of vaginal prolapse, but we could
between the initial indication for hysterectomy and a consecutive confirm a significant impact of heavy work and heavy lifting on the
vaginal prolapse. POP was a significantly more frequent indication for incidence of vaginal prolapse in the women in the study group. Similar
hysterectomy in the study group. There was a statistically significant results were obtained in a previous national survey [22].
difference in the time interval between hysterectomy and vaginal The integrity of pelvic nerves and a possible correlation with pelvic
prolapse repair between women who underwent hysterectomy for floor relaxation is not only influenced by pregnancy and delivery but
POP and women who underwent hysterectomy for other benign also by congenital and postnatal neurological diseases. In the present
pathologies (9 vs 18 years; P b 0.001). study, neurological disability was reported by significantly more
It is probable that women who undergo hysterectomy for POP are women in the study group than in the control group.
more prone to postoperative vaginal prolapse and are also more prone Collagen metabolism undoubtedly plays a major role in the
to its recurrence in a shorter time interval after primary surgery. incidence of POP. A lower synthesis and portion of collagen type 3,
Therefore, it is of paramount importance that the surgeon recognizes impaired metabolism, and a changed proportion of the different types
and corrects the vaginal support defect when performing hysterectomy of collagen influence the laxity and integrity of the endopelvic fascia
to prevent later complications. [23–25] and play a role in the incidence of POP.
Because the problem of POP is vast, there are at least 43 well- Genetic impact on the incidence of vaginal prolapse after hysterec-
documented descriptions of surgical procedures designed to repair tomy was evaluated in the present study by obtaining information from
POP. Effective management of this condition requires knowledge of the patients in both groups about relatives with symptoms of POP. The
anatomy, expertise in accurate physical diagnosis, and skill in the number of relatives with POP symptoms was statistically significantly
30 A. Lukanovič, K. Dražič / International Journal of Gynecology and Obstetrics 110 (2010) 27–30

higher in the women in the study group than in the women in the [5] Mant J, Painter R, Vessey M. Epidemiology of genital prolapse: observations from
the Oxford Family Planning Association Study. Br J Obstet Gynaecol 1997;104(5):
control group. 579–85.
The outcomes of hysterectomy, such as pain and pressure [6] Morley GW, DeLancey JOL. Sacrospinous ligament fixation for eversion of the
sensation, urinary retention, and urinary and fecal incontinence, vagina. Am J Obstet Gynecol 1988;158(4):872–81.
[7] Bump RC, Norton PA. Epidemiology and natural history of pelvic floor dysfunction.
were more frequently reported in the study group women. It is still Obstet Gynecol Clin North Am 1998;25(4):723–46.
unclear whether there is a correlation between the onset and quality [8] Hendrix SL, Clark A, Nygaard I, Aragaki A, Barnabei V, McTiernan A. Pelvic organ
of postoperative wound healing, possible neurovascular damage of prolapse in the Women's Health Initiative: gravity and gravidity. Am J Obstet
Gynecol 2002;186(6):1160–6.
the vaginal supportive structures, and basic surgical technique. In the [9] Brown JS, Waetjen LE, Subak LL, Thom DH, Van den Eeden SK, Vittinghoff E.
present study, postoperative pelvic floor discomfort, pressure, and Pelvic organ prolapse surgery in the United States, 1997. Am J Obstet Gynecol
pain, which predict pelvic floor dysfunction, were present more 2002;186(4):712–6.
[10] Mattingly RF, Thompson JD, editors. Telinde's Operative Gynecology. 6th edn.
often in the study group than in the control group. However, it is not
Philadelphia: J.B. Lippincott; 1985.
known whether the women in the control group did in fact have [11] Birnbaum SJ. Rational therapy for the prolapsed vagina. Am J Obstet Gynecol
pelvic floor discomfort, but did not report the problems to the 1973;115(3):411–9.
healthcare provider. [12] Marchionni M, Bracco GL, Checcucci V, Carabaneanu A, Coccia EM, Mecacci F, et al.
True incidence of vaginal vault prolapse. Thirteen years of experience. J Reprod
The present study shows that in Slovenian women, the risk of Med 1999;44(8):679–84.
vaginal prolapse after hysterectomy is significantly higher if the [13] Swift S, Theofrastous J. Aetiology and classification of pelvic organ prolapse. In:
indication for hysterectomy is POP, if the woman has a positive family Cardozo L, Staskin D, editors. Textbook of Urology and Urogynecology. London:
Isis Medical Media; 2001. p. 576–84.
history of POP, has been doing a substantial amount of heavy work, or [14] Cruikshank SH, Kovac SR. Randomized comparison of three surgical methods used
has had difficult deliveries with birth weight greater than 4000 g. at the time of vaginal hysterectomy to prevent posterior enterocele. Am J Obstet
It is important that before deciding on hysterectomy as the Gynecol 1999;180(4):859–65.
[15] McCall ML. Posterior culdoplasty: surgical correction of enterocele during vaginal
approach to treat a woman with pelvic floor dysfunction, the surgeon hysterectomy: a preliminary report. Obstet Gynecol 1957;10(6):595–602.
should evaluate these risk factors and discuss them with the patient so [16] Wall LL. A technique for modified McCall culdoplasty at the time of abdominal
that she is aware of the complication rates. hysterectomy. J Am Coll Surg 1994;178(5):507–9.
[17] Ostrzenski A. A new, simplified posterior culdoplasty and vaginal vault suspension
during abdominal hysterectomy. Int J Gynecol Obstet 1995;49(1):25–34.
Conflict of interest [18] Cruikshank SH. Preventing posthysterectomy vaginal vault prolapse and enter-
ocele during vaginal hysterectomy. Am J Obstet Gynecol 1987;156(6):1433–40.
[19] Cruikshank SH, Cox DW. Sacrospinous ligament fixation at the time of
The authors have no conflict of interest to declare. transvaginal hysterectomy. Am J Obstet Gynecol 1990;162(6):1611–5.
[20] Brubaker L, Bump R, Jacquetin B, Schuessler B, Weidner A, Zimmern P, et al.
Epidemiology of pelvic organ prolapse. In: Abrams P, Cardozo L, Khoury S, Wein A,
References editors. Incontinence: Proceedings of the 2nd International Consultation on
Incontinence. 2nd edn. Plymouth: Health Publication Ltd; 2002. p. 243–65.
[1] Abrams P, Cardozo L, Fall M, Griffiths D, Rosier P, Ulmsten U, et al. The standardisation [21] Samuelsson EC, Victor FT, Tibblin G, Svardsudd KF. Signs of genital prolapse in a
of terminology of lower urinary tract function; report from the Standardisation Sub- Swedish population of women 20 to 59 years of age and possible related factors.
Committee of the International Continence Society. Neurourol Urodyn 2002;21(2): Am J Obstet Gynecol 1999;180(2 Pt1):299–305.
167–78. [22] Bai SW. The role of collagen formation in pelvic floor disorder. Gynecol Forum
[2] De Lancey JOL. Functional anatomy of the pelvic floor and urinary continence 2004;9(1):10–2.
mechanism. In: Schussler B, Laycock J, Norton P, Stanton S, editors. Pelvic Floor Re- [23] Sustersic O, Kralj B. The influence of obesity, constitution and physical work on the
education. Principles and Practice. London: Springer Verlag; 1994. p. 9–21. phenomenon of urinary incontinence in women. Int Urogynecol J Pelvic Floor
[3] Stovall TG. Hysterectomy. In: Berek JS, editor. Novak's Gynecology. 13th edn. Dysfunct 1998;9(3):140–4.
Philadelphia: Lippincott Williams & Wilkins; 2002. p. 761–801. [24] Gill EJ, Hurt WG. Pathophysiology of pelvic organ prolapse. Obstet Gynecol Clin
[4] Olsen AL, Smith VJ, Bergstrom JO, Colling JC, Clark AL. Epidemiology of surgically North Am 1998;25(4):757–69.
managed pelvic organ prolapse and urinary incontinence. Obstet Gynecol 1997;89(4): [25] Norton PA, Baker JE, Sharp HC, Warenski JC. Genitourinary prolapse and joint
501–6. hypermobility in women. Obstet Gynecol 1995;85(2):225–8.