INTERVIEW-BASED OR SELF-ADMINISTEREDST. MARK'S INCONTINENCE SCORE, - IS THERE A DIFFERENCE?
H. H. JOHANNESSEN1,S. NORDERVAL 2, A. STORDAHL 1, A. WIBE 3;
1Ostfold Hosp. Trust, Sarpsborg, Norway, 2Departmentof Surgery, Univ. Hosp. of North-Norway, Tromsø, Norway, 3NorwegianUniv. of Sci. and Technology, Trondheim, Norway.
Introduction: The St. Mark’sincontinence score (SMIS) measures the frequency and severity of analincontinence (AI) symptoms, and was originally designed and validatedfor use in an interview setting (iSMIS). There is conflictingevidence of the validity of self-administered SMIS(sSMIS).
Objective: To compare self-administered andinterview-based reports of AI symptoms made on the same day using theSMIS.
Methods: Postpartum women reported symptoms of AI ona sSMIS in the patients’ waiting area at their local anorectaloutpatients’ clinic prior to clinical investigations. Afterclinical investigations, an iSMIS was completed by one of twoconsultant surgeons who were blinded to the self-administered reportsof AI. Assessment of correlation and agreement between the individualitems of the iSMIS and the sSMIS was assessed using Spearman’s rhoand weighted kappa statistics, respectively.
Results: Atotal of 147 women were included approximately one year postpartum,and the mean age was 30.1 years. The mean iSMIS and sSMIS reportedwas 4.0 (SD:3.6) and 4.3 (SD:4.0), respectively. Spearman’s rhoshowed a strong relationship between the two total SMIS scores(r=.769, n=147, p<.001), and explained variance was 59% (r²=.591).Except for in the individual item 1c; “incontinence of gas”,women reported more frequent AI symptoms on the sSMIS compared to theiSMIS. The assessment of consistency between the individual items ofthe iSMIS and sSMIS showed substantial agreement (κ≥0.60) for allitems except fair agreement for item 1a;“incontinence of formedstool” (κ=0.22) , and moderate for item 1d;“change in lifestyle”(κ=0.5)(Table 1) (Landis & Koch, 1977).
Table 1.Agreement between interview-based (iSMIS) and self-reported symptoms(sSMIS) of anal incontinence using the St. Mark’s incontinencescore (n=147 paired assessments).
Exactagreement / agreement ± 1 point (%)
*Weighted Kappa; squared weights;1.000, 0.9375, 0.7500, 0.4375
Conclusions: Although theSt. Mark’s incontinence score was originally designed to measurethe degree of anal incontinence in an interview setting, it seemslike more severe incontinence was reported when the patients filledin the forms themselves. As shown in previous studies (Bartlett etal, 2007; Guise et al, 2007), it may be that in this population ofyoung, healthy postpartum women, embarrassment plays a role whenreporting frequency of incontinence symptoms in an interview setting.Few women reported weekly or daily incontinence of formed stool,which may have affected the agreement between the interview-based andself-administered incontinence score in this particular item.Considering the level of correlation and agreement between the twomethods of reporting anal incontinence symptoms, the St. Mark’sscore may be used as both an interview-based as well as aself-administered incontinence score.
References: BartlettL, Nowak M, Ho YH. Reasons for non-disclosure of faecal incontinence:A comparison between two survey methods. Tech Coloproctol. 2007Sep;11(3):251-7. Guise JM, Morris C, Osterweil P, Li H, Rosenberg D,Greenlick M. Incidence of fecal incontinence after childbirth. ObstetGynecol. 2007 Feb;109(2 Pt 1):281-8. Landis JR, Koch GG. Themeasurement of observer agreement for categorical data. Biometrics.1977:159-74.