STATISTICAL ANALYSIS PLAN. Addendum 1 (V 1 12SEP2015)

Size: px

Start display at page:

Download "STATISTICAL ANALYSIS PLAN. Addendum 1 (V 1 12SEP2015)"

Ada Bell
6 years ago
Views:

1 Pelvic Flr Disrder STATISTICAL ANALYSIS PLAN Addendum 1 (V 1 12SEP2015) PFDN Prtcl Number 22P01: Refractry Overactive Bladder: Sacral NEurmdulatin v. BTulinum Txin Assessment (ROSETTA) AUTHOR (S): Tracy Nlen Dennis Wallace Versin V 1.0, 12SEP2015; Page i Dwnladed Frm: n 02/04/2018

2 Pelvic Flr Disrder 1 MODIFICATIONS AND CLARIFICATIONS TO PLANNED ANALYSES 1.1 Bladder Diaries Tw primary types f activities have ccurred as part f QA/QC prcesses fr the ROSETTA bladder diary data that have impacted hw we will analyze the study data. The first set f activities include a central audit f bladder diaries and the secnd being a set f visits t the Kaiser-Dwney (Bellflwer) site. During the centralized audit f bladder diaries, it was identified that: Surce diaries culd nt be identified fr participant 25RS047 at 4 and 5 mnths. Participant 07RS054 nly had 2 valid days f incntinence episde data reprted at baseline. The first site visit t Kaiser was cnducted Octber 6-7, During the visit a number f ptential issues regarding data quality at the site were identified. Based n these cncerns, a secnd, fr cause, site visit was scheduled t the Bellflwer facility during Nvember During that site visit, all available study participant diaries frm all randmized ESTEEM and ROSETTA participants were reviewed. Key findings that impact analysis apprach decisin fr ROSETTA included: Surce diaries culd nt be fund fr 23 fllw-up and 7 baseline visits. After the visit, a subset f these diaries were lcated and determined t be valid diaries. The final set f missing diaries include: 20RS001: 9 mnths 20RS007: 2-12 mnths 20RS009: 2 and 12 mnths 20RS011: 3 mnths 20RS014: 1 mnth 20RS015: 1 mnth 20RS018: Baseline 20RS019: Baseline 20RS024: Baseline 20RS025: Baseline Fllw-up fr all Interstim participants were cmpleted using 7 day diaries instead f 3 day diaries, a clear prtcl deviatin that can bias study results Participant 20RS023 was determined t be ineligible at baseline based n crrectins made t extracting data frm the baseline diary. Based n these findings the fllwing decisins regarding analyses were made: Data withut a surce diary will be excluded frm the primary bladder diary analyses. Participants with missing baseline diaries will be cmpletely excluded frm these analyses as change frm baseline utcmes cannt be calculated reliably at any visit. Sensitivity analyses that summarize the change in summary statistics fr the bladder diary endpints when Versin V 1.0, 12SEP2015; Page 2 Dwnladed Frm: n 02/04/2018

3 Pelvic Flr Disrder including and excluding these diaries will be dne t assess the impact f the exclusin f these diaries n study results. Participants identified during the audit r site visits as being ineligible at baseline based n a review f diaries will be excluded frm all study analyses f efficacy. If treated, they will remain in the safety ppulatin. Baseline diaries fund during the audit r site visits t nly have tw valid days fr randmized participants will be cnsidered valid fr the purpses f identifying eligibility as well as fr use in study analyses. 1.2 Masked Data Review Findings After the database was lcked fr the 6 mnth primary utcme analyses but prir t analysis f any study data, the prtcl team reviewed key data in a masked fashin (e.g. with dummy IDs and n infrmatin abut treatment assignment where pssible) t make final decisins regarding analysis ppulatin membership as well as t address any utstanding questins with respect t which data shuld be included in the analyses. The final decisins are belw: Analysis Ppulatin Membership: Safety, ITT and Clinical Respnder Ppulatins: Per the SAP, individuals nt receiving full treatment n Interstim are excluded frm the CR ppulatin. Althugh nt mentined in the SAP, we are similarly excluding Btx individuals that did nt receive a cmplete 1 st injectin. In reviewing the prtcl deviatins fr the PP ppulatin, a deviatin was nted that the reprted respnse diary fr a participant was invalid. Participant 24RS045 will be classified as a nnrespnder and be excluded frm the CR ppulatin. PP Ppulatin: UI Therapies: A DCC nurse crdinatr reviewed all reprted therapies t ensure that all urinary incntinence (UI) therapies were apprpriately identified. The study PI reviewed all UI therapies. Any data cllected after the use f the fllwing medicatins will be excluded frm PP analyses (antichlinergics and beta-3 adrenergic agnist specifically will be summarized fr the primary manuscript and are identified in red fnt belw): 200 UNITS BOTOX INJECTION INTERSTIM IMPLANTATION. OXYBUTININ PATCH 2ND INTERSTIM PLACEMENT INTERSTIM PLACEMENT OXYBUTYNIN BOTOX INTERSTIM PLACEMENT (STAGE I AND II) OXYBUTYNIN (ONLY BEFORE STUDY TREATMENT AND DIARY) BOTOX INJECTION INTERSTIM STAGE 1 IMPLANTED OXYBUTYNIN(DITROPAN) BOTOX INJECTION (IN BLADDER) INTERSTIM STAGE 2 IMPLANTATION OXYTROL PATCH BOTOX INJECTION 200U INTERSTIM STAGE 2 PLACEMENT PERCUTANEOUS TIBIAL NERVE STIMULATION (PTNS). CROSSOVER TO INTERSTIM DEVICE INTERSTIM STAGE 2 SURGERY PTNS SESSIONS DARIFENACIN (ENABLEX) MIRABEGRON PTNS SESSIONS. DETROL MIRABEGRON 25MG 1 PO EVERYDAY SANCTURA Versin V 1.0, 12SEP2015; Page 3 Dwnladed Frm: n 02/04/2018

4 Pelvic Flr Disrder DETROL LA MIRABEGRON 50 MG 1 DAILY SANCTURA 1 PO BID DETROPAN XL MIRABEGRON E.R. SOLFENACIN (VESICARE) DITROPAN MIRABETRIQ- 25MG SOLIFENACIN (VESICARE). DITROPAN XL. MIRABEYRON TOLTERODINE ENABLEX MIREBEGRON TOLTERODINE (DETROL LA) FSLP MYBETRIQ TOVIAC FSLP/INTERSTIM MYBETRIQ 50 MG 1 EVERY OTHER DAY TOVIAZ GELNIQUE MYRBETRIQ TOVIAZ 8MG PO ER 1 PO DAILY GELNIQUE, APPLY 1 PK OF GEL TO SKIN DAILY MYRBETRIQ - 50MG TOVIAZ 8MG PO QD INTERSTIM MYRBETRIQ -25MG TROPIUM (SANCTURA) INTERSTIM 1 MYRBETRIQ 25MG TROSPIUM INTERSTIM 2 ONABOTULINUMTOXINA (BOTOX) TROSPIUM (SANCTURA XR) INTERSTIM BATTERY AND LEAD WIRE REPLACED OXYBURIN GHS TROSPIUM (SANCTURA) INTERSTIM DEVICE OXYBUTANIN TROSPIUM (SANCTURA). INTERSTIM DEVICE STAGE I OXYBUTIN VESICARE Althugh reprted as UI therapies, data will nt be excluded after the use f the fllwing treatments: ACCUPUNCTURE CAPSAICIN.025 % EX CREAM COAPTITE INJECTION DESIPRAMINE (NORPRAMIN) ESTRONG VAGINAL RING IMIPRAMINE KEGELS (SELF-ADMINSTERED) MID-URETHRAL SLING NORTRIPTYLINE PELVIC FLOOR PHYSICAL THERAPY PESSARY PREMARIN Interstim Device Usage: Visits with the Interstim device turned n <50% f the time since last visit are t be excluded frm PP analyses. There are sme visits where device use was nt cllected. Versin V 1.0, 12SEP2015; Page 4 Dwnladed Frm: n 02/04/2018

5 Pelvic Flr Disrder Visits where device use infrmatin was nt available due t a missed visit r prgrammer nt being brught in will use device use % frm the next attended visit if ne exists. Otherwise device use will be assumed t be 0%. Fr data thrugh 6 mnths n participants nt excluding frm the per-prtcl ppulatin fr ther reasns: Missed visits r visits with n prgrammer with n subsequent cmpleted visit: 25RS032: Mnths 3-6; 26RS024: 3-6. Missed visits r visits with n prgrammer with subsequent cmpleted visit: 02RS052- Mnths 1-3 use; 02RS063- Mnths 3-6; 02RS066- Mnths 1-3; 07RS044- Mnths 0-1; 08RS014- Mnths 0-1; 15RS016- Mnths 1-3; 15RS024- Mnths 1-6; 20RS009- Mnths 0-1; 23RS051- Mnths 3-6; 24RS033- Mnths 1-3; 24RS036- Mnths 1-3; 24RS041- Mnths 3-6 (visit cmpleted but participant did nt have time t d prgramming assessment but site nted everything wrking fine); 25RS021- Mnths 0-1 (visit cmpleted but participant did nt have time t d prgramming assessment but site nted everything wrking fine); 25RS032- Mnths- 0-1; 25RS048- Mnths 1-3. Visits with n dcumented reasn fr why prgramming infrmatin was nt reprted will be assumed t be 0%. Fr data thrugh 6 mnths n participants nt excluding frm the per-prtcl ppulatin fr ther reasns: 02RS050- Mnths 0-1 use; 07RS031- Mnths 1-3; 24RS030- Mnths 0-1. Visits with n prgramming infrmatin due t technical difficulty will be assumed t be 0% as we are unsure if device was reset. Fr data thrugh 6 mnths n participants nt excluding frm the per-prtcl ppulatin fr ther reasns: 08RS008- Mnths 0-1 use (device wuld nt sync) 14RS012- Mnths 0-1 use (prgrammer device brken) 25RS050- Mnths 3-6 use (malfunctin f prgrammer) 25RS052- Mnths 1-6 use (prgrammer nt brught t 3 mnth and then malfunctined at 6 mnth) 25RS053- Mnths 3-6 use (malfunctin f device) 25RS054- Mnths 0-3 use (prgrammer nt brught t 1 mnth visit and then printer didn t wrk at 3 mnth) Prtcl Deviatins: The study PI reviewed all eligibility, randmizatin and treatment administratin deviatins as well as any deviatins that mentined key wrds (e.g. Interstim, Btx, injectin, implantatin) t identify any additinal deviatins that wuld excluded all r part f a participant s data frm the PP analyses. Likewise, the DCC study team reviewed all ther deviatins. The fllwing decisins were made: All data after a participant receives an incmplete r nn-prtcl dse re-injectin will be excluded frm PP analyses (e.g. Subject 25RS009). Study therapy deviatins related t safety r peratinal requirements that shuld nt impact the therapy efficacy will nt have data excluded as a result (e.g. btx prcedure Versin V 1.0, 12SEP2015; Page 5 Dwnladed Frm: n 02/04/2018

6 Pelvic Flr Disrder in an individual that was culture psitive, Subject 08RS003 (excluded fr ther reasn: incmplete injectin); <2 prgrams determined during FSLP test perid, Subject 08RS011; different gauge needle r dilutin amunt used than that specified in prtcl, subject 16RS014 and 24RS015) Data Inclusin Decisins: Diary Participants with delays in treatment (r decisin t nt treat) mre than 90 days frm randmizatin were t have their baseline diary repeated. Diaries were nt repeated fr 27 participants. Amng thse ultimately treated, all participants were treated within 180 days f cmpleting their baseline diary except fr ne individual wh was treated at 302 days pst-diary cmpletin. The prtcl team reviewed lags between baseline diaries and treatment initiatin and determined that all baseline diaries wuld be included in the analysis and n participants wuld be excluded frm any analysis ppulatin based n the timing f their baseline diary. Diaries nt cmpleted n actual frms will be included in analyses. Althugh the MOP specified fllw-up diaries must have 2 cnsecutive days t be cnsidered valid, any recrded diaries with valid data fr at least 2 days regardless f if cnsecutive will be analyzed. As detailed in the SAP, utcme data are analyzed based actual study mnth within which they were btained as ppsed t reprted study mnth. If there are n utcme data within a visit windw, then data utside the visit windw may be used. Clarificatins t the analysis plan: Out f windw utcme data will nly be used fr a particular visit if it des NOT fall int the windw f anther visit. Visit windws were +/- 10 days thrugh 6 mnths and +/-28 days. Fr analysis, the fllwing rules will be emplyed regarding visit windws: Data mre than 5 days utside f the windw thrugh 6 mnth will nt be cnsidered fr analysis. This implies an verall +/-15 day windw thrugh 6 mnths and excludes very little data as mst windws are nearly back t back. Data mre than 14 days utside f the windw fr visits after 6 mnths will nt be cnsidered fr analysis implying an verall +/- 42 day windw. Site specific plts f mean and median diary data utcmes were reviewed (e.g. site median f mean UUIE/day at each fllw-up visit). The fllwing bservatins were explred. N additinal items were identified fr fllw-up. Stress Episdes: Overall the mean numbers are lw which suggests we did identify ur ppulatin f interest. N site trends appear cncerning. Urge Episdes: One site has higher visit means cmpared t the ther sites; hwever, the site s medians d nt differ frm ther sites. There d appear t be mre subjects that cntinue t have high values pst-baseline at this site than ther sites; hwever, a single extreme subject seems t be the primary cntributr t the site means differing. Other Episdes: One site has higher visit means cmpared t the ther sites; hwever, the site s medians d nt differ frm ther sites. A single extreme subject seems t be the primary cntributr t the site means differing. Day Vids and Pad Use: N site trends appear cncerning. Versin V 1.0, 12SEP2015; Page 6 Dwnladed Frm: n 02/04/2018

7 Pelvic Flr Disrder Night Vids: One site has higher visit means and medians cmpared t the ther sites. There d appear t be mre subjects that cntinue t have high values pst-baseline at this site than ther sites Data Inclusin Decisins: QOL Participants with delays in treatment (r decisin t nt treat) mre than 90 days frm randmizatin had NO requirements fr repeating baseline QOL assessments. QOL calls ccurred mre than 90 days prir t treatment r decisin t nt treat fr 22 participants and OABq-SF was cmpleted mre than 90 days prir fr 53 participants. Amng thse ultimately treated, all participants were treated within 180 days f cmpleting their baseline QOL call except fr tw individuals wh were treated at 234 and 269 days pst-call cmpletin. Amng thse ultimately treated, all participants were treated within 180 days f cmpleting their baseline OABq-SF except fr 4 individuals wh were treated at 204, 221, 283, and 288 days pst-oabq- SF cmpletin. The prtcl team reviewed lags between baseline QOL assessments and treatment initiatin and determined that all baseline QOL assessments wuld be included in the analysis and n participants wuld be excluded frm any analysis ppulatin based n the timing f their baseline QOL assessment. QOLs filled ut with assistance r n paper will be included in all analyses. As detailed in the SAP, utcme data are analyzed based actual study mnth within which they were btained as ppsed t reprted study mnth. If there are n utcme data within a visit windw, then data utside the visit windw may be used. Clarificatins t the analysis plan: Out f windw utcme data will nly be used fr a particular visit if it des NOT fall int the windw f anther visit. Visit windws were +/- 10 days thrugh 6 mnths and +/-28 days. Fr analysis, the fllwing rules will be emplyed regarding visit windws: Fr QOL calls ccurring n mre frequently than 6 mnths apart: Data mre than 32 days utside f the windw at 6 mnth will nt be cnsidered fr analysis. This implies an verall +/-42 day windw at 6 mnths Data mre than 14 days utside f the windw fr visits after 6 mnths will nt be cnsidered fr analysis implying an verall +/- 42 day windw. Fr OABq-SF, the same rules defined fr diaries abve will be emplyed as this assessment was dne n the same time table as diaries. Site specific plts f mean and median QOL data utcmes were reviewed. N site trends have been identified f cncern. N additinal items are identified fr fllw-up Data Inclusin Decisins: Other Pints Incmplete r invalid UDS at baseline will nt preclude an individual frm analyses except fr thse that use UDS as a ptential predictr. Per prtcl and MOP, the definitin f a UTI event f interest is symptmatic, culture + (r culture culd nt be perfrmed) and treated. Since sites generally tk a brader reprting apprach, we have reviewed and marked whether each event meets the prtcl definitin. Versin V 1.0, 12SEP2015; Page 7 Dwnladed Frm: n 02/04/2018

8 Pelvic Flr Disrder Detailed ntes are in the UTIS (MaskedReview_UTIs_wMaskedNtes.dcx) but generally, if the prtcl definitin culd nt be cnfirmed as met (e.g. incnclusive culture r nt symptms specified) then the event is nt cunted as a UTI. AEs f pain, shrtness f breath r muscle weakness AEs that may pssibly represent pain, shrtness f breath r muscle weakness were identified as being f interest fr primary manuscript. These events will be summarized by SOC and Preferred Term t determine apprpriate classificatin fr manuscript. 1.3 QOL Scring Updates OAB-SATq The scring manual fr the OAB-SATq states that higher scres shuld crrespnd t better utcmes and als details that final cmpsite scres fr each dimensin shuld be btained by subtracting raw scres frm the highest pssible scre and dividing by range. Hwever, based n the directin f scre fr individual respnses, this derivatin wuld result in lwer scres crrespnding with better utcmes. We have adjusting the scring algrithm instead t be: 100*(raw scre lwest pssible scre)/scre range. This calculatin results in higher scres crrespnding t better utcmes and an verall scale f Versin V 1.0, 12SEP2015; Page 8 Dwnladed Frm: n 02/04/2018

The estimator, X, is unbiased and, if one assumes that the variance of X7 is constant from week to week, then the variance of X7 is given by

The estimator, X, is unbiased and, if one assumes that the variance of X7 is constant from week to week, then the variance of X7 is given by ESTIMATION PROCEDURES USED TO PRODUCE WEEKLY FLU STATISTICS FROM THE HEALTH INTERVIEW SURVEY James T. Massey, Gail S. Pe, Walt R. Simmns Natinal Center fr Health Statistics. INTRODUCTION In April 97, the