Pregnancy and delivery: a urodynamic viewpoint

Transcription

1 British Journal of Obstetrics and Gynaecology November 2000, Vol107, pp Pregnancy and delivery: a urodynamic viewpoint *C. Chaliha Research Fellow (Urogynaecology),** J. M. Bland Professor (Medical Statistics),?A. Monga Consultant (Gynaecology),*Ss. L. Stanton Professor (Urogynaecology), $A. H. Sultan Consultant (Gynaecology and Obstetrics) YJrogynaecology Unit and **Department of Public Health Sciences, St. George S Hospital Medical School, London;?Princess Anne Hospital, Southampton; +May Day University Hospital, Thornton Heath, Surrey Objective The aims of this study were to establish prospectively the prevalence of objective bladder dysfunction before and after delivery by means of urodynamic investigations and to assess the effect of obstetric variables on bladder function. Design Prospective longitudinal study, Twin channel subtracted cystometry was performed in the standing and sitting position, with a cough stress test at the end of filling. The investigations were repeated three months postpartum. Participants Two hundred and eighty-six nulliparae with singleton pregnancies who were delivered between April 1996 and November 1997 attended for antenatal assessment after 34 weeks of gestation and 161 who returned postpartum. Setting Department of Obstetrics and Gynaecology in a London teaching hospital. Results The mean urodynamic values both in pregnancy and postpartum lower than values defined in a non-pregnant population. The prevalence of genuine stress incontinence and detrusor instability were antenatally 9% and 8%, respectively, and postpartum 5% and 7%, respectively. Obstetric and neonatal factors were not related to urodynamic variables. Conclusions Despite the reported high prevalence of urinary incontinence related to pregnancy and childbirth, neither pregnancy nor delivery resulted in any consistent effects on objective bladder function. Postpartum urodynamic measurements were not related to either obstetric or neonatal variables, but were dependent on antenatal values. INTRODUCTION There is a strong association between pregnancy and delivery and the development of urinary symptoms; however the pathophysiology is poorly understood'-3. Urinary incontinence may result from a combination of factors, including pudendal nerve damage, muscle trauma and changes in urethral support4". However, most women have recovered nerve function by two months postpartum, despite persistence of symptoms, so the significance of nerve damage is unclear4. Alterations in urethral support occur in more than 50% of women after vaginal delivery, and this is thought to be a consequence of decreased pelvic floor ~trength~.~. Changes in collagen strength and structure in pregnancy'.'', and an individual predisposition to incontinence due to congenital alterations in collagen type may also contribute to the development of urinary symptoms". Whether it is pregnancy itself or delivery that initiates these changes Correspondence: Dr C. Chaliha, Urogynaecology Unit, Department of Obstetrics and Gynaecology, St. George's Hospital, Cranmer Terrace, London SW 17 ORE, UK. is not known. Urodynamic assessment of bladder function in pregnancy has demonstrated the occurrence of detrusor instability and genuine stress incontinence, particularly antenatally'2-14. Although many studies have reported urinary symptoms in relation to pregnancy and delivery, there are few studies assessing the effect of pregnancy and childbirth on urodynamic variables. The aims of this study were to establish prospectively the prevalence of objective bladder dysfunction in a cohort of pregnant women before and after delivery using urodynamic investigations. The women in the study were nulliparae with no previous history of bladder injury, surgery or congenital abnormalities, and who were therefore considered ideal participants in the assessments of the effects of pregnancy and delivery on bladder function. The findings were related to obstetric variables to identify possible risk factors for objective bladder dysfunction. METHODS Nulliparous women attending the antenatal clinics of St George's Hospital between April 1996 and August RCOG 2000 British Journal of Obstetrics and Gynaecology

2 URODYNAMICS IN PREGNANCY AND DELIVERY 1355 were invited to participate in this study. All women were evaluated after 34 weeks of gestation and asked to attend for repeat investigations 12 weeks after delivery. Exclusion criteria included pre-existing diabetes mellitus, neurological disorders, a history of urinary tract surgery or anatomical abnormalities, and active urinary tract infection. The women were asked to attend with a full bladder, which was then catheterised with a filling 12-French gauge Nelaton catheter and a 3.5-French gauge nylon catheter to measure the intravesical pressure. The volume was drained and measured. A 4-French gauge catheter with a perforated balloon at the tip of the transducer was inserted rectally. Detrusor pressure was estimated by subtracting the abdominal pressure from the intravesical pressure. Adequate subtraction was then checked by asking the patient to cough, and filling was started in the sitting position. The bladder was filled with 0.9% normal saline at room temperature at a rate of 100 ml/min. The woman was asked to inform the investigator when she reached various stages of bladder sensation (i.e. the first sensation of bladder filling, usually ml; the sensation of a strong desire to void; and at maximum capacity when the bladder felt completely full, usually ml. During the filling phase the woman was asked to cough and stimuli such as running water were introduced to provoke detrusor activity. A change in position from sitting to standing was used to assess detrusor activity. The bladder was emptied once cystometric capacity was reached and the test was repeated with the woman standing throughout. After cystometric capacity had been reached, the filling catheter was removed and the woman was asked to cough while standing, Any abnormal detrusor activity or incontinence was noted. The woman then voided into a flowmeter so that the maximum voiding pressure, peak flow rate and volume voided could be measured. Methods, definitions and units of measurement conformed to the standards of the International Continence Society, unless otherwise statedts. Pregnancy and delivery details were obtained from the women s case records after the postnatal visit. All the urodynamic investigations were performed by the same investigator (C.C.) who was unaware of the details of the delivery. Statistical analysis All the information was entered into the Access database; Stata 5.0 (Stata Corp, College Station, Texas) was used for analysis. The urodynamic variables all had positively skewed distributions. Logarithmic transformation improved the approximation to the Normal considerably (Fig. l), although some were over-transformed and became negatively skewed. Despite this, the log transformed data could all be regarded as a reasonable approximation to a Normal distribution, and so these variables were analysed on the log scale using two-sample t tests, correlation coefficients and linear regression. Differences between groups and their confidence intervals were antilogged to give ratio estimates. Explanatory variables were restricted to the antenatal variables of interest: mode of delivery as a threelevel factor (e.g. normal vaginal delivery, assisted delivery, caesarean section); the length of the passive phase of the second stage of labour (which may affect urodynamic variables by distension of the vagina and pressure on the bladder and urethra); the length of the active phase of the second stage of labour (which may affect urodynamic variables by injury to the vagina, bladder and urethra); epidural analgesia (which may interfere with voiding of the bladder during and immediately after labour and so affect urodynamic variables); catheterisation in labour which may affect urodynamic variables by exciting detrusor contractions; perineal trauma; augmentation of labour; fetal weight; and fetal head circumference. The duration of second stage of labour was also positively skewed and so log transformed. Fig. 1. Second desire to void standing, on natural (a) and logarithmic (b) scales.

3 1356 C. CHALIHA ET AL. This study was approved by the Research and Ethics Committee of St George s Hospital and all women gave informed consent. RESULTS Of the 286 women who attended for antenatal investigations, 161 returned postpartum. In those women who attended for antenatal investigations the mean age (range) was 29 years (17-43). The delivery details of the women who attended for urodynamic investigations, compared with the hospital nulliparous booking population delivered over the same time period, are represented in Table 1. Contractions in labour were augmented using a standard regimen of intravenous oxytocin if progress was felt to be slow. Approximately 60% of women had epidural analgesia, the remainder receiving nitrous oxide, the transcutaneous electrical nerve stimulation (TENS) machine or pethidine. If vaginal assisted delivery was required and epidural analgesia was not in use, local anaesthetic infiltration or a pudendal block was administered. All episiotomies were mediolateral. Four women sustained third degree tears, three of whom returned for postnatal follow up. Women were catheterised before an assisted vaginal delivery; the other women who had a catheter in labour required this because of voiding difficulties in the first stage. The mean birthweight (standard deviation) of the infants was 3360 g (480), and the mean (SD) head circumference was 35.2 cm (4-2). These data were compared with nulliparae delivered in the hospital over the same time (Table 1). Antenatally, the prevalence of genuine stress incontinence and of detrusor instability were 9% and 8%, respectively; postpartum the prevalence was 5% and 7%, respectively. The mean urodynamic values in the third trimester and postpartum were lower than normal limits defined irl a non-pregnant population. Comparison of antenatal and postnatal measurements showed a statistically significant increase in first sensation and in strong sensation to void on both sit and stand fill phases, an increase in maximum cystometric capacity on standing and a decrease in peak flow rate (Table 2). If we adjust these P values by the Bonferroni correction to allow for multiple testing, multiplying each P value by 8, these remain highly significant, showing that overall there is a clear change in urodynamic variables following delivery. Analysis of urodynamic measurements by mode of delivery showed no evidence of any difference between women who had a normal vaginal delivery, instrumental delivery or caesarean section (Table 3). Table 4 shows correlation coefficients between the continuous obstetric variables (e.g. duration of labour and fetal measurements). Of 32 comparisons, two were significant at the 5% level only; because of the number of comparisons we can discount these and conclude that there is no evidence that these variables were related to postpartum urodynamic measurements. Table 4 also shows the correlations with the corresponding antenatal values of the measurement of interest. These were mostly statistically significant and positive, though not large. The relationship of postpartum urodynamic measurements to dichotomous obstetric variables is shown in Table 5. Only augmentation of labour showed any evidence of a relationship with postnatal urodynamics. If we apply the stringent Bonferroni criterion and multiply the P values by 8, these too cease to be significant (P = 0-06). We Table 1. Characteristics of the women in the study who attended for antenatal and postnatal investigations compared with other nulliparae delivered in the hospital. Values are given as n (%) or mean [SD]. NA = not available. Modes of delivery Study population (n = 161) Hospital nulliparous population (n = 2040) Spontaneous 89 (55.3) 1133 (55.5) Forceps/ventouse 41 (25.5) 456 (22.4) Caesarean section (total) 31 (19.3) 451 (22.1) Prelabour caesarean section 9 (5.6) 86 (4.2) Caesarean section - 1st stage of labour 22 (13.7) 365 (17.9) Augmented labour 59/152 (38.8) 913/1954 (46.7) Epidural analgesia in labour 84/152 (55.3) 822/1954 (42.1) Perineal laceration 65/130 (50) 834/1589 (52.5) Episiotomy 38/130 (29.2) 671/1589 (42.2) Third degree tear 3/130 (2.3) 28/1589 (1.8) Catheter in labour 69/130 (53.1) 796/1589 (50.1) Length of 1st stage (min) 581 [349.2] NA Length of passive 2nd stage (min) 35.6 [31.2] NA Length of active 2nd stage (min) 52.9 [33.4] NA 1 st urodynamic test (gestational age) 36.7 [ nd urodynamic test (postnatal weeks) 14.3 [2.3] -

4 URODYNAMICS IN PREGNANCY AND DELIVERY 1357 Table 2. Urodynamic variables before and after delivery in women who returned for postnatal investigations. Values are given as mean (SD), unless otherwise indicated. FDV = first desire to void; SDV = strong desire to void; MXCC = maximum cystometric capacity; MVP = maximum voiding pressure; P A = postnatal : antenatal. Urodynamic variable Antenatal values (n = 161) Postnatal values (n = 161) PA ratio 95% CI for ratio* P* Sit SDV (d) Stand SDV (ml) MVP(cm H,O) Peak flow rate (mlls) iii (70-1) 188 (91.3) 301 (146.1) 96 (69.3) 167 (102.0) 239 (136.4) (n = 125) 38 (22.0) (n = 153) 28 (16.3) 148 (83.0) 217 (94.3) 299 (114.4) 139 (90.2) 209 (105.0) 271 (123.6) (n = 118) 32 (17.2) (n = 157) 23 (14.4) < 0~ < 0~0001 < 0~0001 0~ *Paired t method on log transformed data. Table 3. Urodynamic measurements according to mode of delivery. Values are given as mean (SD), unless otherwise indicated. FDV = first desire to void; SDV = strong desire to void; MXCC = maximum cystometric capacity; MVP = maximum voiding pressure. Normal vaginal Instrumental Caesarean section Mode of delivery delivery (n = 89) delivery (n = 41) (n = 31) P* Sit SDV (ml) Stand SDV (ml) MVP (cm H,O) Peak flow rate ( d s ) 150 (83) 225 (96) 300 (114) 96 (68) 200 (93) 264 (14) (n = 70) 33 (19) (n = 85) 24 (18) 138 (87) 203 (96) 309 (126) 162 (115) 231 (131) 290 (146) (n = 27) 30 (13) (n = 38) 22 (8) 153 (80) 214 (87) 281 (101) 140 (75) 205 (98) 265 (119) (n = 28) 33 (17) (n = 30) 23 (9) *F test on log-transformed data, adjusted for antenatal values. Table 4. Correlation coefficients between postnatal urodynamic measurements, the corresponding antenatal measurements, and obstetric variables. Values are given as [ 1. FDV = first desire to void; SDV = strong desire to void; MXCC = maximum cystometric capacity; MVP = maximum voiding pressure; LPS = length of passive second stage; U S = length of active second stage. Postnatal urodynamic Corresponding antenatal Fetal Fetal head variable vaiable LPS LAS weight circumference 0.18* Sit SDV (ml) 0.23' ' 0.21* ' Stand SDV (ml) 0.36' ' MVP (cm H,O) * Peak flow rate ( U s ) 0.17* Tests of significance using log-transformed data: *P < 0.05, +P < 0.01, *P < 0,001. continued the analysis by adjusting these differences for the antenatal measurement (Table 6), which produced little change in the results. Thus we can conclude that the only obstetric variable for which we have any evidence of an effect on urodynamic variables postpartum is augmentation of labour, and here the evidence is weak. DISCUSSION This is the largest prospective study using urodynamic measurements to assess bladder function objectively in pregnancy and postpartum. Although there was good evidence for changes in urodynamic variables following

5 1358 C. CHALIHA ET AL. Table 5. Postnatal urodynamic measurements and obstetric variables. Ratios of means for those with characteristic compared with those without. Values are given as ratio (P). FDV = first desire to void; SDV = strong desire to void; MXCC = maximum cystometric capacity; MVP s maximum voiding pressure; LPS = length of passive second stage; LAS = length of active second stage. Urodynamic variable Epidural Perineal trauma Augmentation of labour Catheter Sit SDV (ml) Stand SDV (mlj MVP (cm H,O) Peak flow rate (mvs) 1.05 (0.6) 1 *OO (0.9) 0.99 (0.9) 1.12 (0.3) 1.03 (0.7) 1-04 (0.6) 1.09 (0.4) 1-06 (0.5) 0.92 (0.5) 0.97 (0.6) 0.99 (0.8) 0.88 (0.2) 0.94 (0.5) 0.94 (0.4) 1.01 (0.9) 0.88 (0.1) 1.27 (0.03) 1.22 (0.008) 1.11 (0.1) 1.33 (0.01) 1.16 (0.07) 1.19 (0.02) 1.11 (0.4) 1.07 (0.4) 0.91 (0.4) 0.92 (0.2) 0.94 (0.3) 1.12 (0.3) 0.97 (0.7) 0.97 (0.7) 0.91 (0.4) 0.98 (0.8) delivery, we could find little evidence that any particular features of delivery influenced this. Urinary incontinence is a common symptom, with up to 85% of women reporting incontinence in pregnancyl6.l7 and up to 35% postpartum''. Urodynamics are a standard investigation to assess bladder dysfunction and we wished to see whether the high prevalence of symptoms would be reflected in objective measurement of bladder function. Muellner'' performed 86 cystometrograms in pregnancy and reported an increase in bladder capacity to an average of 1300 ml in the third trimester, associated with bladder hypotonia. This finding was disputed by Francis' who found no change in bladder capacity in the first trimester, and a reduced bladder capacity in the third trimester in association with increased detrusor irritability rather than bladder hypotonia. Our findings suggest that bladder capacity shows no evidence of hypotonia in the third trimester and postpartum. The lower bladder capacity in pregnancy may be caused by pressure of the gravid uterus on the bladder. Our data agree with Kerr-Wilson et alz0 who found that all urodynamic variables were lower than normal limits at 48 hours and 4 weeks postpartum. The values of Table 6. Effect of augmentation of labour on urodynamic variables, adjusted for antenatal measurement. FDV = first desire to void; SDV = strong desire to void; MXCC = maximum cystometric capacity; MVP = maximum voiding pressure. Urodynamic variable Ratio* P 95% CI Sit SDV (ml) Stand SDV (ml) MVP (cm H,O) Peak flow rate (mlls) *Augmented compared with nonaugmented labours to to to to to to to to 1.26 first sensation and maximum cystometric capacity were lower than in a nonpregnant population, and this may account for the symptoms of sensory urgency, frequency and nocturia experienced in pregnancy. Bladder function itself was not significantly affected by delivery, as there was no evidence of a relationship between urodynamic measurements and obstetric and neonatal variables. This is at variance with symptom studies that have shown a significant relationship between postpartum stress incontinence and vaginal delivery. This difference is either because urodynamics are not an appropriate method to assess bladder function in pregnancy or because the pathophysiology of postpartum urinary incontinence is more strongly related to structural changes in the urinary tract and loss of bladder neck support than intrinsic bladder function. Other diagnostic tests that may be considered are penned ultrasound, magnetic resonance imaging and urethral pressure profilometry. Alternatively, it may be possible that changes in bladder function related to pregnancy occur before the third trimester, and so would not have been revealed in this study. Therefore, despite the high prevalence of urinary symptoms in pregnancy and postpartum, urodynamic investigations did not reveal a consistent effect on bladder function. The changes in bladder function are consistent with the pressure effects of the gravid uterus and do not appear to be related to the method of delivery or birthweight. Further studies are required to understand the pathophysiology of bladder disorders in pregnancy and postpartum. Acknowledgements The authors would like to thank the South Thames Project Grant Scheme for funding this study. References 1 Francis WJA. Disturbances of bladder function in relation to pregnancy. J Obstet Gynaecol Br Emp 1960; 61: Beck RP, Hsu N. Pregnancy, childbirth and the menopause related to the development of stress incontinence. Am J Obstet Gynecoll965; 91:

6 URODYNAMICS IN PREGNANCY AND DELIVERY Stanton SL, Ken-Wilson R, Hams GV. The incidence of urological symptoms in normal pregnancy. Br J Obsrer Gynaecol 1980; 87: Snooks SJ, Swash M, Setchell M, Henry MM. Injury to the innervation of pelvic floor sphincter musculature in childbirth. Lancer 1984; 2: Snooks SJ, Badenoch OF, Tiptaft RC, Swash M. Perineal nerve damage in genuine stress incontinence. An electrophysiological study. Br J Uroll985; 57: Allen RE, Hosker GL, Smith ARB, Warrell DW. Pelvic floor damage and childbirth: a neurophysiological study. Br J Obster Gynaecol 1990; 97: Sampselle CM. Changes in pelvic floor strength associated with childbirth. J Obstet Gynecol Neonatal Nurs 1990; 19: Peschers U, Schaer G, Anthuber C, Delancey JOL, Schuessler B. Changes in vesical neck mobility following vaginal delivery. Obstet Cgnecol1996; 88: Landon CR, Crofts CE, Smith ARB. Mechanical properties of fascia during pregnancy: a possible factor in the development of stress incontinence of urine. Conremp Rev Obsrer Gynaecoll990; Lavin JM, Smith ARB, Anderson J et al. The effect of the first pregnancy on the connective tissue of the rectus sheath. Neurourol Umdyn 1997; 16: Norton P, Boyd C, Deak S. Collagen synthesis in women with genital prolapse or stress urinary incontinence. Neurourol Urodyn 1992; 11: Clow WM. Effect of posture on the bladder and urethral function in normal pregnancy. Urol Inr 1975; 30: Cutner A, Cardozo LD, Benness CJ. Assessment of urinary symptoms in early pregnancy. BrJObster Gynaecoll991; 98: Cutner A, Cardozo LD, Benness CJ. Assessment of urinary symptoms in the second half of pregnancy. Int Urogynecol J 1992; 3: Abrams P, Blaivas JG, Stanton SL, Andersen JT. The standardisation of terminology of lower urinary tract function. Scund J Urol Nephml 1988; 114 (SUPPI): Francis WJA. The onset of stress incontinence. J Obstet Gynaecol Br Emp 1960; 67: Viktrup L, Lose G, Rolff M, Barfoed K. The symptom of stress incontinence caused by pregnancy or delivery in primiparas. Obsret Gynecoll992; 79: Wilson PD, Herbison RM, Herbison GP. Obstetric practice and the prevalence of urinary incontinence three months after delivery. Br J Obstet Gynaecoll996; 103: Muellner SR. Physiological bladder changes during pregnancy and the puerperium. J Urol 1939; 41: Ken-Wilson RHJ, Thompson SW, Orr JW, Davis RO, Cloud GA. Effect of labor on the post-partum bladder. Obster Gynecol 1984; 64: Accepted 7 June 2000