Appendix G. References for Fault Tree Model Probability Estimates

Proactive Risk Assessment of Surgical Site Infection in Ambulatory Surgery Centers

IdentifierDescriptionInput ValueReference
Event115Poor hygiene while dressing0.1Best guess
Event116Linen contaminated0.05Best guess
Event136Fail to wash hands properly (OR staff)0.12Sharir et al. 2001—literature indicates this probability is in general much lower (around 12%) for OR staff doing invasive procedures than other staff (where it can be as high as 40%)
Event138Fail to remove watch/jewelry/fake nails0.15Kennedy et al. 2004—this is a conservative estimate, can be as high as 30-40%
Event142Staff not well-trained in infection control0.3Babcock et al. 2002 based on multi-institutional survey, but highly depends on the site; sensitivity analysis needed
Event145Skin lesion near incision site0.001Prevalence of skin lesion estimated at 0.002, split equally  btw Ev 145 and 147
Event146Fail to halt opern prior to identifying skin lesion0.001Best guess (sub-TEP indicated the probability of this failure was small)
Event147Open skin lesions (entire body)0.001Prevalence of skin lesion estimated at 0.002, split equally  btw Ev 145m 147
Event148Fail to schedule OP after infections are resolved0.001Best guess (sub-TEP indicated the probability of this failure was small)
Event169Other pt SSI risk0.00046Calculated from other probability estimates in the model (1—proportion of all other patient types connected to Gate 2)
Event173Staff fails to provide pt with instruction (for weight reduction)0.9From Tony (not done in ASCs at  all, so 0.9 is a conservative estimate, implying this is done 10% of the time)
Event182Fail to administer indicated antibiotics0.2Burke et al. 2001, Silver et al. 1996., Van Kasteren et al. 2007 (high variation in estimates, ranging from 20% to 60% — sensitivity analysis req'd)
Event192Fail to use correct antiseptic0.003Gt 236=0.125 (fail to prep appropriately, from Babcock et al. 2003), split equally among Ev 192, 193, 613, 614. So = 0.125/4=0.003125
Event193Fail to use for appropriate time0.003Gt 236=0.125 (fail to prep appropriately, from Babcock et al. 2003), split equally among Ev 192, 193, 613, 614. So = 0.125/4=0.003125
Event203Fail to wash hand (non-OR staff)0.38Sharir et al. 2001
Event21Pt fails to follow instruction (for smoking cessation)0.13From Tony, proportion of pts who do not disclose smoking status is 50% (agrees with the literature; Schofield et al indicates this is 40%). For smokers who disclose their status (50% of them), 13-15% will not follow instructions for smoking cessation.
Event210Lack of 2 filter beds in series0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split among Ev 210, 612, 213, 611, 325, 328, 326, 327
Event213Not introducing air at ceiling0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split among Ev 210, 612, 213, 611, 325, 328, 326, 327
Event220For visible soiling, fail to use disinfection aft OP0.000033From Tony, Fail to clean OR=0.0001 (very small probability), divided equally among Ev 220, 221, and 222.
Event221Poor routine cleaning after each OP0.000033From Tony, Fail to clean OR=0.0001 (very small probability), divide equally among Ev 220, 221, and 222.
Event222Fail to wet-vacuum floor w/ disinftn after last OP0.000033From Tony, Fail to clean OR=0.0001 (very small probability), divide equally among Ev 220, 221, and 222.
Event235Drape gets contamtd during opern0.001667From Tony, Draping-related failure=0.01, split equally among its children events. So 0.01/6=0.001667
Event236Fail to notice drape not sterilize0.001667From Tony, Draping-related failure=0.01, split equally among its children events. So 0.01/6=0.001667
Event237Drape pulls bug to SS0.001667From Tony, Draping-related failure=0.01, split equally among its children events. So 0.01/6=0.001667
Event239Fail to prevent hypothermia0.01From Tony
Event240Inappropriate use of drain material0.01From Tony
Event241Tourniquet time >60min0.082Demers et al. 1998
Event242Excessive surg time0.05Best guess
Event270Anesthetic risk0.05DeLee. 1985
Event28Pt colonized with MRSA0.1Kim et al. 2010, Barbos et al. 2010., large variation, from 4.4% to 37.2%, sensitivity analysis req'd.
Event29Non-MRSA pt (complement to Gt 28)0.9Complement to Ev 28, 1-0.1=0.9
Event30Fail to prep skin appropriately0.125Babcock et al. 2003, Gottrup et al. 2005, variation from 12.5% to 25%
Event306Not assessed during pre-op visit (obesity)0.01Best guess
Event307Not assessed during pre-op call (obesity)0.05From Tony
Event322Fail to notice drape has holes0.001667From Tony, Draping-related failure=0.01, split equally among its children events. So 0.01/6=0.001667
Event323Break in steril technique/field0.02From Tony
Event325Temp ctrl problems0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split btw Ev 210, 612, 213, 611, 325, 328, 326, 327
Event326Humidity problems0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split btw Ev 210, 612, 213, 611, 325, 328, 326, 327
Event327Air flow problems0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split btw Ev 210, 612, 213, 611, 325, 328, 326, 327
Event328Other (ventilation not working prblems)0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split btw Ev 210, 612, 213, 611, 325, 328, 326, 327
Event329Surgeon fails to inform staff of special needs0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event330Other (fail to min entry/exit and inappr traff pattns)0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event331Lunch breaks0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event332Other (fail to minimize staff changes)0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event333Fail to restrict residents0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event334Fail to restrict family members0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event336Not enough instruments or instrument failure0.003DeLee. 1985.
Event34Fail to shave SS right before procedure0.01From Tony; similar % in Mangram et al., 1999.
Event353Bug moves to pt (fail to decontaminate bug from staff)0.15Best guess
Event373Bug moves to pt (MRSA or other bugs on the stethoscopes)0.05Best guess
Event38Not noticed or managed0.001Best guess
Event393Fail to closure appropriately0.001From Tony
Event394Bug gets to SS (fail to closure appropriately)0.15Best guess
Event395Bug gets to SS (fail to suture appropriately)0.15Best guess
Event396Staff contaminated0.05Kim et al. 2010.
Event404Poor post-op directions to pt0.15From Jackson et al. 1991., for 68% of these patients, the mean reading comprehension level was below the ninth grade while most of the commonly used brochures provided to the patients were written at a level far above their ability to comprehend (77% of the materials were written at the eleventh-grade level). Kincey et al. 1975 estimates this as 0.15.
Event405Pt understands but not follows directions0.1Mangram et al. 1999 and Czarneki et al. 2007.
Event417Pt not available for post-op call0.15From Tony
Event418Nurse fails to recognize infection0.12Nurse not assesses 0.16*(1-Event 417)=0.12
Event419Pt fails to come for post-op visit0.15From Tony
Event421Post-op visit but physicn fails to manage infect0.01Mangram et al. 1999
Event422Fails to notice or mang (pt gets bug during post-op visit)0.2Best guess
Event423Pt gets bug during post-op visit0.1Best guess
Event424Bug moves to SS (pt gets bug dur post-op vst and not mang)0.01Best guess
Event433Pt has bug post-op (pt has bug after post-op visit)0.05Best guess
Event45Inadequate draping0.001667From Tony, Draping-related failure=0.01, split equally among its children events. So 0.01/6=0.001667.
Event450Obese but not diabetes pt (30≤BMI<40)0.29333.8% obese (not morbidly obese) in the US adult population. Then Prevalence of obese pt -Ev653 (obese and diabetic)=0.338-0.045=0.293.
Event456Pt is a smoker and has COPD0.053395(proportion of smokers-proportion of elderly smokers)*prevalence of COPD in the smoking population. Adult Cigarette Smoking in US: 20.6% (from CDC). 13.62% patients are elderly (≥65 years of age) in the arthroscopic knee surgery population [CA 2008 SASD]. From Stang et al. 2000., 30% of smokers have COPD. So (0.206-0.1362*0.206)*0.30=0.05338284.
Event459SSI if bug gets to SS (elderly, not diabetic, and non-smoker)0.014 Kaye et al. 2005
Event46Fail to airdry disinf bfr drape0.001667From Tony, Draping-related failure=0.01, split equally among its children events. So 0.01/6=0.001667
Event47Fail to suture appropriately0.001From Tony
Event473Pt w/ vascular compromise in limb0.01From Shammas 2007., the prevalence of peripheral vascular disease in the general population is around 10%, so estimated the prevalence of pt w/ vascular compromise in limb as 1%
Event475SSI if bug gets to SS (pt w/ skin lesion, near incision site and fail to mitigate)0.03Best guess
Event476SSI if bug gets to SS (pt w/ skin lesion, entire body and fail to mitigate)0.03Best guess
Event48Antiseptic not used0.1Mangram et al. 1999
Event482Staff fails to provide pt with instruction (for compromised immune system patients)0.05Best guess
Event483Pt fails to follow instruction (for compromised immune system patients)0.05Best guess
Event49Site not cleansed appropriately0.1Best guess
Event494AIDS pt0.004Marks et al. 2006, prevalence of AIDS patients in the US
Event496Pt with systemic steroid use0.01Babcock et al. 2003
Event497Pt w/ rheumatoid disease0.05Grennan et al. 2001
Event498Cured or not under treatment0.000031CA 2008 SASD Database, proportion of cancer patients undergoing arthroscopic surgery used as an estimate of this probability
Event499SSI if bug gets to SS (cancer pt cured or not under treatment)0.01Berbari et al. 1998
Event50Dressing torn off0.01From Tony
Event500SSI if bug gets to SS (AIDS pt, risk not mitigated)0.09Abalo et al. 2010
Event501Risk mitigated (complement to Gt 249)0.8gate 24
Event502SSI if bug gets to SS (AIDS pt, risk mitigated)0.06Abalo et al. 2010
Event503SSI if bug gets to SS (pt w/ rheumatoid disease)0.0187Best guess
Event504Risk mitigated (complement to Gt 249)0.8Complement to Gt 249, so 1-0.2=0.8
Event505SSI if bug gets to SS(pt w/ rheumatoid disease SSI risk, mitigated)0.01Best guess
Event506SSI if bug gets to SS (pt w/ Systemic steroid use, risk not mitigated)0.0187Post et al. 1993
Event507Risk mitigated (complement to Gt 249)0.8Complement to Gt 249, so 1-0.2=0.8
Event508SSI if bug gets to SS (pt w/ Systemic steroid use, risk mitigated)0.0187Assume if mitigated, then risk same as healthy patient
Event509Pt on Immunosuppressive drugs0.001Prevalence in the US is 300,000/300M=0.1%
Event51Dressing not applied correctly0.01From Tony
Event510SSI if bug gets to SS (pt on Immunosuppressive drugs)0.0187Best guess
Event512Risk mitigated (complement to Gt 249)0.8Complement to Gt 249, so 1-0.2=0.8
Event513SSI if bug gets to SS (Pt on Immunosuppressive drugs, mitigated)0.01Best guess
Event516Fail to assess during vitals check0.0005Best guess (very small)
Event517Fail to reschedule OP (after high temp assessed)0.0005Best guess (very small)
Event519Glucose level controlled pre-op (Gt 10 complement)0.97Gt 10 complement
Event52IV phlebitis0.016Arjona et al. 1997
Event520Glucose level not controlled post-op0.27Latham et al. 2001
Event521SSI if bug gets to SS (diabetes (not obese) pt, GL controlled pre-op but not post-op)0.03Boulton et al. 2005.
Event522SSI if bug gets to SS (diabetes (not obese) pt, GL controlled post-op but not pre-op)0.03Boulton et al. 2005.
Event523Glucose level controlled post-op0.73Latham et al. 2001
Event524Glucose level controlled pre-op0.97same as Event 519
Event525SSI if bug gets to pt (diabetes  (not obese) pt,GL controlled, pre-op and post-op)0.02Boulton et al. 2005.
Event526Glucose level controlled post-op0.73same as Event 523
Event527Glucose level not controlled post-op0.27same as Event 520
Event528SSI if bug gets to pt (diabetes (not obese) pt, GL fail to control pre-op and post op)0.077Latham et al. 2001
Event529Malnourished pt0.001From Tony (very small proportion of malnourished pts in the US)
Event53Pt hands contaminated0.05Best guess
Event530Nutritional status not improved pre-op1From Tony (ASCs do not assess nutritional status)
Event531SSI if bug gets to SS (malnourished pt, not mitigated)0.05Klein et al. 1996
Event539Alcoholic0.04Spies et al. 2004 (proportion of alcoholism in the US population)
Event543SSI if bug gets to SS (obese pt, weight not redcd and nutrition not impr)0.0325Olsen et al. 2008
Event544Weight  reduced and nutrition impr0.1Best guess (not routinely done in ASCs for obese pts)
Event545SSI if bug gets to SS (obese pt, weight reduced and nutrition impr)0.01Best guess
Event546Pt does not provide accurate info0.37From Schofield et al. 1999, smoking self-disclosure rate=0.63, so 1-0.63=0.37
Event550Acute risk0.03from Myles et al. 2002, smoker SSI risk split equally btw Ev 550, 558
Event552Stp smk pre-op but starts in rec phase0.2Complement of Gt 14, 0.5*0.4=0.2. From Evangelista et al. 2000., compliance rates range from 10%-85% depending on the population and the medical regimen studied. Estimated that pt lack of compliance w/ discharge orders=40%
Event553Acute risk0.03Myles et al. 2002, smoker SSI risk split equally btw Ev 553, 554
Event554Chronic risk (underlying heart disease)0.03Myles et al. 2002, smoker SSI risk split equally btw Ev 553, 554
Event555Pt stops smoking pre-op and post-op0.3(1-Gt 14)*pt complies post-op=(1-0.5)*0.6=0.3
Event558Chronic risk (underlying heart disease)0.03Myles et al. 2002, smoker SSI risk split equally btw Ev 550, 558
Event561SSI if bug gets to SS (pt w/ vascular compromise in limb)0.02Best guess
Event591Bug moves to SS (due to inappropriate prep)0.692From Kim et al. 2010, for MRSA patients
Event592Bug moves to SS (due to inappropriate prep)0.1Best guess
Event593Staff w/ active skin infection0.0525From Jeans et al. 2010., rate of S. aureus carriage in health-care workers: 10.5%-39.7%, conservative estimate of 10.5% split btw Ev 593, 594
Event594Other (rather than staff w/ active skin infection)0.0525From Jeans et al. 2010., rate of S. aureus carriage in health-care workers: 10.5%-39.7%, conservative estimate of 10.5% split btw Ev 593, 594
Event599Fail to report infection0.01Mangram et al. 1999
Event605Infection subtle0.04Best guess
Event610Inappropriate traffic patterns0.014286Gt 64 High traffic in OR=0.1 (Pryor et al., 1998.), equally split btw Ev 329, 610, 330, 331, 332, 333, 334
Event611Improper air exchange0.00625Gt 105 improper ventilation system=0.05 (best guess), equally split btw Ev 210, 612, 213, 611, 325, 328, 326, 327
Event612Insufficient efficiency0.00625Gt 105 improper ventilation system=0.05 (Pryor et al., 1998.), equally split btw Ev 210, 612, 213, 611, 325, 328, 326, 327
Event613Contaminating the site0.003Gt 236=0.125 (fail to prep appropriately, from Babcock et al. 2003), split equally among Ev 192, 193, 613, 614. So=0.125/4=0.003125
Event614Cleansing from dirty to clean0.003Gt 236=0.125 (fail to prep appropriately, from Babcock et al. 2003), split equally among Ev 192, 193, 613, 614. So=0.125/4=0.003125
Event618SSI if bug gets to SS (alcoholic pt, mitigated)0.01Best guess
Event619SSI if bug gets to SS (alcoholic pt, not mitigated)0.026Spies et al.2004.
Event621Pt does not provide accurate info (on alcoholic status)0.37Schofield et al. 1999., 63% of patients revealed to admission staff that they were alcoholic, so 1-0.63=0.37
Event622Pt fails to comply with instruction (on stopping drinking—alcoholic pt)0.13Event 21 estimate (for smokers) used here also
Event624Mitigated (for alcoholic pt)0.5From Tony
Event625Glucose level not controlled pre-op0.03From Tony
Event627Fail to assess, ctrl pre-op and postpone surg (vascular compromise in limb)1From Tony
Event628GLl not controlled pre-op (Ev 519 complement)0.03Event 519 complement, 1-0.97=0.03
Event630Acute and chronic risk (pt stops smoking pre-op and post-op)0.01Best guess
Event631Fail to assess compromised immu sys0.1Best guess
Event632Dialysis pt0.001As of 2007 there were 368,544 US residents with end stage renal disease receiving dialysis. (Kidney diseases), estimated as 0.001
Event633Fail to screen pre-op (for MRSA) (dialysis pt)1Routine MRSA test not done—site visits
Event634Fail to screen pre-op (for MRSA) (elderly pt)1Routine MRSA test not done -site visits
Event635Elderly pt0.1362From CA 2008 SASD
Event636Fail to screen pre-op (for MRSA) (nursing home pt)1Routine MRSA test not done-site visits
Event637Nursing home pt0.005Jones et al. 2009., as of 2004 there were 1,492,200 nursing home patients in the USA, estimated as 0.005
Event638Fail to screen pre-op (for MRSA) (pt with history of MRSA)1Routine MRSA test not done-site visits
Event639Pt with history of MRSA0.002Klein et al. 2007. The estimated number of MRSA hospitalizations in 2005 was 278,203, estimated as 0.002
Event640Fail to screen pre-op (for MRSA) (other pt)1Routine MRSA test not done-site visits
Event641Other pt0.85581-all the non "other pts", such as dialysis pt, etc.
Event642Bug gets to pt (fail to protect pt effectively)0.2Best guess
Event643Bug gets to pt (surgical procedure puts pt at higher risk)0.2Best guess
Event644Fail to readmin (antibiotics) for longer surgery0.1Burke et al. 2001., Silver et al. 1996., Van Kasteren et al. 2007.
Event645Morbidly obese pt (BMI>=40)0.057CDC web site and Flegal et al. 2002.
Event647Elderly, diabetic, and non-smoker0.020644Elderly pt prevalence*percentage of elderly diabetic pt*(1-prevalence of smokers). Elderly pt prevalence= 13.62%  [CA 2008 SASD arthroscopic knee surgery patient population]. Percentage of elderly diabetic pt = 19% [http://www.cdc.gov/diabetes/statistics/prev/national/figbyage.htm]. Prevalence of smokers in the US =20.6% (CDC). So 0.1362*0.19*(1-0.206)=0.0206
Event648Elderly, not diabetic, and non-smoker0.08734Elderly pt prevalence*(1-percentage of elderly diabetic pt)*(1-prevalence of smokers). Elderly pt prevalence= 13.62% [CA 2008 SASD arthroscopic knee surgery patient population]. Percentage of elderly diabetic pt = 19% [http://www.cdc.gov/diabetes/statistics/prev/national/figbyage.htm].  Prevalence of smokers in the US =20.6% (CDC). So 0.1362*(1-0.19)*(1-0.206)=0.087
Event649SSI if bug gets to SS (elderly, diabetic, and non-smoker)0.0378Olsen et al. 2002.
Event651Diabetes pt not diagnosed0.026It is estimated that 23.6 million people are afflicted in the United States, accounting for  approximately 7.8% of the population. Nearly one third of persons with diabetes are unaware that they have DM. Boulton et al. 2005.
Event652SSI if bug gets to SS (diabetes pt not diagnosed)0.02Boulton et al. 2005.
Event653Diabetes and obese pt0.04538.6% of the obese suffer diabetes [http://usgovinfo.about.com/library/weekly/aa010803a.htm]. Prevalence of obese pt who are diabetics around 4.5%
Event654Diabetes but not obese pt0.007From McMillen 2011, nearly one third of persons with diabetes are unaware that they have DM. So two thirds of the diabetes pt-diabetes and obese pt=0.078*(2/3)-0.045=0.007
Event655SSI if bug gets to SS (diabetes (and obese) pt, GL controlled pre-op but not post-op)0.077Olsen et al. 2002.
Event656SSI if bug gets to SS (diabetes (and obese) pt, GL controlled post-op but not pre-op)0.077Olsen et al. 2002.
Event657SSI if bug gets to pt (diabetes  (and obese) pt,GL controlled, pre-op and post-op)0.04Olsen et al. 2008
Event658SSI if bug gets to pt (diabetes (and obese) pt, GL fail to control pre-op and post op)0.1Boulton et al. 2005.
Event659Glove puncture0.18Duke Infection Control Group 2011.
Event660Pt fails to notice infection0.16Mangram et al. 1999
Event670Elderly, diabetic,  and smoker0.005356Elderly pt prevalence *percentage of elderly diabetic pt* prevalence of smokers. Elderly pt prevalence= 13.62% [CA 2008 SASD arthroscopic knee surgery patient population]. Percentage of elderly diabetic pt = 19% [http://www.cdc.gov/diabetes/statistics/prev/national/figbyage.htm]. Prevalence of smokers in the US=20.6% (CDC). So 0.1362*0.19*0.206=0.0053
Event671SSI if bug gets to SS (elderly, diabetic, and smoker)0.1134Olsen et al. 2002.
Event672Elderly, not diabetic, and smoker0.02266Elderly pt prevalence*(1-percentage of elderly diabetic pt)*prevalence of smokers. Elderly pt prevalence= 13.62% [CA 2008 SASD arthroscopic knee surgery patient population]. Percentage of elderly diabetic pt = 19% [http://www.cdc.gov/diabetes/statistics/prev/national/figbyage.htm]. Prevalence of smokers in the US =20.6% (CDC). So 0.1362*(1-0.19)*0.206=0.02
Event673SSI if bug gets to SS (smoker and elderly, not diabetic)0.042Kaye et al. 2005.
Event674MRSA or other bugs on the stethoscopes0.32Merlin et al. 2009
Event676SSI if bug gets to SS (malnourished pt, mitigated)0.01Assume, when mitigated, risk same as healthy pt
Event677Pt is a smoker but has no COPD0.124589(prevalence of smokers-elderly smokers)*(1-prevalence of COPD in the smoking population). Adult Cigarette Smoking in US: 20.6% (CDC). % elderly patients (>=65 years of age) in the arthroscopic knee surgery population is 13.62% [CA 2008 SASD arthroscopic knee surgery patient population]. From Stang et al. 2000., 30% of smokers have COPD. So (0.206-0.1362*0.206)*(1-0.30)=0.12458
Event678Stp smk pre-op but starts in rec phase0.2Same as Ev552
Event679Pt stops smoking pre-op and post-op0.3Same as Ev555
Event680Acute and chronic risk (pt stops smoking pre-op and post-op)0.01Assume, when mitigated, risk same as health pt
Event682Pt fails to follow instruction0.13Same as Ev21
Event683Acute risk (COPD)0.06Best guess
Event684Chronic risk (underlying heart disease, COPD)0.06Best guess
Event685Acute risk (COPD)0.06Best guess
Event686Chronic risk (underlying heart disease, COPD)0.06Best guess
Event687Pt does not provide accurate info0.37Same as Event 621
Page last reviewed April 2013
Internet Citation: Appendix G. References for Fault Tree Model Probability Estimates : Proactive Risk Assessment of Surgical Site Infection in Ambulatory Surgery Centers. April 2013. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/research/findings/final-reports/stpra/stpraapg.html