Using implementation science to decrease variation and high opioid administration in a surgical ICU.

BACKGROUND: High doses and prolonged duration of opioids are associated with tolerance, dependence, and increased mortality. Unfortunately, despite recent efforts to curb outpatient opioid prescribing because of the ongoing epidemic, utilization remains high in the intensive care setting, with intubated patients commonly receiving infusions with a potency much higher than doses required to achieve pain control. We attempted to use implementation science techniques to monitor and reduce excessive opioid prescribing in ventilated patients in our Surgical ICU. METHODS: We conducted a prospective study investigating opioid administration in a closed SICU at an academic medical center over 18 months. Commonly accepted conversions were used to aggregate daily patient opioid use. Patients with a history of chronic opioid use and those being treated with an ICP monitor/drain, neuromuscular blocker, or ECMO were excluded. If the patient spent a portion of a day on a ventilator, that day's total was included in the "vent group." MMEs per patient were collected for each patient and assigned to the on-call intensivist. Intensivists were blinded to the data for the first seven months. They were then provided with academic detailing followed by audit & feedback over the subsequent 11 months, demonstrating how opioid utilization during their time in the SICU compared to the unit average and a blinded list of the other attendings. Student's T-tests were performed to compare opioid utilization before and after initiation of academic detailing and audit & feedback. RESULTS: Opioid utilization in patients on a ventilator decreased by 20.1% during the feedback period, including less variation among all intensivists and a 30.9% reduction by the highest prescribers. CONCLUSION: Implementation science approaches can effectively reduce variation in opioid prescribing, especially for high outliers in a SICU. These interventions may reduce the risks associated with prolonged use of high-dose opioids. LEVEL OF EVIDENCE: Prospective pre-post-intervention, Level II.

Perinatal HIV Prevention Outcomes in U.S.-Born Versus Foreign-Born Blacks, PSD Cohort, 1995-2004.

We examined differences in HIV-infected U.S.-born and foreign-born black mothers who delivered perinatally HIV-exposed and -infected children during 1995-2004 in the Pediatric Spectrum of HIV Disease Project, a longitudinal cohort study. Prevalence ratios were calculated to explain differences in perinatal HIV prevention opportunities comparing U.S.-born to foreign-born and African-born to Caribbean-born black mothers. U.S.-born compared with foreign-born HIV-infected black mothers were significantly more likely to have used cocaine or other non-intravenous illicit drugs, exchanged money or drugs for sex, known their HIV status before giving birth, received intrapartum antiretroviral (ARV) prophylaxis, and delivered a premature infant; and were significantly less likely to have received prenatal care or delivered an HIV-infected infant. African-born compared with Caribbean-born black mothers were more likely to receive intrapartum ARV prophylaxis. These differences by maternal geographical origin have important implications for perinatal HIV transmission prevention, and highlight the validity of disaggregating data by racial/ethnic subgroups.

Repeat axillofemoral grafting as treatment for axillofemoral graft occlusion.

BACKGROUND: Patency of failed axillofemoral (ax-fem) grafts following thrombectomy is so poor, aortofemoral grafts are recommended as treatment for ax-fem graft thrombosis. In patients who are not candidates for aortic grafting, repeat ax-fem grafting is an alternative to thrombectomy. This report compares our experience treating ax-fem graft thrombosis with replacement or revision vs thrombectomy. METHODS: Patients treated with ax-fem grafts from October 1985 to April 2001 were identified, and those who underwent reoperation for thrombosis were reviewed. Limb salvage and patency of revision procedures (thrombectomy vs repeat ax-fem grafting) were determined using Kaplan-Meier curves. RESULTS: Three hundred thirty-five patients underwent ax-fem grafting, and 39 (11.6%) of the 335 required reoperation for graft failure. Twenty-five of these 39 patients had 51 operations for graft thrombosis: 42 graft replacements and/or anastomotic revision(s), and 9 thrombectomies. At 18 months, mean +/- SD patency following thrombectomy was 11% +/- 10%, while that for graft replacement or anastomotic revision was 54% +/- 8% (P<.001). Limb salvage at 18 months following revision for thrombosis was 88% +/- 5%. CONCLUSIONS: The large majority of ax-fem grafts do not require reoperation. For failure due to thrombosis, repeat ax-fem grafting provides excellent limb salvage. Axillofemoral graft replacement and/or anastomotic revision has superior patency to thrombectomy.

Carotid surgery following previous carotid endarterectomy is safe and effective.

With the perceived high risk of repeat carotid surgery, carotid angioplasty and stenting have been advocated recently as the preferred treatment of recurrent carotid disease following carotid endarterectomy. An experience with the operative treatment of recurrent carotid disease to document the risks and benefits of this procedure is presented. A review of a prospectively acquired vascular registry over a 10-year period (Jan. 1990-Jan. 2000) was undertaken to identify patients undergoing repeat carotid surgery following previous carotid endarterectomy. All patients were treated with repeat carotid endarterectomy, carotid interposition graft, or subclavian-carotid bypass. The perioperative stroke and death rate, operative complications, life-table freedom from stroke, and rates of recurrent stenosis were documented. During the study period 56 patients underwent repeat carotid surgery, comprising 6% of all carotid operations during this period. The indication for operation was symptomatic disease recurrence in 41 cases (73%) and asymptomatic recurrent stenosis >/=80% in 15 cases (27%). The average interval from the prior carotid endarterectomy to the repeat operation was 78 months (range 3 weeks-297 months). The operations performed included repeat carotid endarterectomy with patch angioplasty in 31 cases (55%), interposition grafts in 19 cases (34%), and subclavian-carotid bypass in 6 cases (11%). There were three perioperative strokes with one resulting in death for a perioperative stroke and death rate of 5.4%. One minor transient cranial nerve (CN IX) injury occurred. Mean follow-up was 29 months (range, 1-116 months). Life-table freedom from stroke was 95% at 1 year and 90% at 5 years. Recurrent stenosis (>/=80%) developed in three patients (5.4%) during follow-up, including one internal carotid artery occlusion. Two patients (3.6%) underwent repeat surgery. Repeat surgery for recurrent cerebrovascular disease following carotid endarterectomy is safe and provides durable freedom from stroke. Most patients are candidates for repeat endarterectomy with patching, but interposition grafting is often required. These results strongly support the continued role of repeat carotid surgery in the treatment of recurrent carotid disease.

Chronic traumatic popliteal arteriovenous fistula: Case report.

Chronic popliteal arteriovenous fistula is a common result of penetrating vascular trauma. Such a case, involving a young, active-duty soldier who received a gunshot wound to the back of his left leg, is reported. Operative findings confirmed the presence of a fistula and a dilated, tortuous popliteal vein. A 1-cm communication between the main popliteal vein and artery just above the knee was found. Division, with lateral venorrhaphy and end-to-end repair of the artery, was performed, with good results. Particularly stressed in this report is the importance of increased physician awareness, combined with an aggressive surgical approach to reduce the morbidity, especially the high amputation rate so often associated with this kind of penetrating vascular trauma.

Risk factors for autogenous infrainguinal bypass occlusion in patients with prosthetic inflow grafts.

OBJECTIVE: In patients with prosthetic inflow (PI) grafts the proximal anastomosis of autogenous infrainguinal bypass (AIB) can be placed on the PI or on a distal native vessel in the groin. This study was performed to determine the effect of placement of an AIB proximal anastomotic site in a patient with ipsilateral PI. METHODS: Patients undergoing AIB and PI between January 1990 and July 2002 were included in the study. They were classified into two groups on the basis of location of the proximal anastomosis. In group 1 the AIB proximal anastomosis was placed on the PI in the groin, whereas in group 2 the AIB proximal anastomosis was placed on a distal native groin artery. Patency, limb salvage, and patient survival in the two groups were calculated with the Kaplan-Meier method. The Cox proportional hazards model was used to determine independent risk factors affecting AIB patency. RESULTS: Two hundred twenty-nine patients underwent AIB and PI. In group 1, 23 AIBs became thrombosed concurrent with 26 PI occlusions, and in group 2, 7 AIBs became thrombosed concurrent with 36 PI occlusions (P <.001). Five-year assisted primary patency, limb salvage, and patient survival in groups 1 and 2 were 50% and 75% (P <.001, log-rank test), 78% and 90% (P =.005, log-rank test), and 56% and 69% (P = NS, log-rank test), respectively. Factors independently associated with AIB occlusion are hypertension (hazard ratio [HR], 3.41; 95% confidence interval [CI], 1.65-7.05; P =.001), postoperative warfarin sodium therapy (HR, 1.86; 95% CI, 1.07-3.23; P =.03), continued smoking (HR, 1.72; 95% CI, 0.93-3.18; P =.08), AIB arising from PI (HR, 2.38; 95% CI, 1.35-4.18; P =.003), and PI occlusion (HR, 3.70; 95% CI, 2.15-6.36; P <.001). CONCLUSION: A proximal AIB anastomosis located directly on the PI is an independent risk factor for decreased AIB patency of equal or greater importance than current smoking, hypertension, or PI occlusion. The proximal anastomosis of an AIB in a patient with an ipsilateral PI should be placed on a distal native artery.

Comparison of procedural outcomes after lower extremity reversed vein grafting and secondary surgical revision.

OBJECTIVE: Many lower extremity vein graft procedures require revision. Although morbidity associated with revision procedures is assumed minimal, this has not been previously quantified and may be underestimated. In this study, patient outcome after initial vein graft procedures and revisions are compared. METHODS: Records for all patients undergoing vein graft revision from January 1995 to August 2002 were reviewed for operation time, estimated blood loss, blood transfusion, hospital length of stay, perioperative complications, and functional status at discharge and at 2-month follow-up. Revisions were compared with the original operation and by revision type. RESULTS: One hundred sixty-five vein graft revisions were performed in 137 patients. In comparison with the initial bypass procedure, mean operation time (3.35 +/- 1.41 hours vs 2.58 +/- 1.04 hours; P <.001), estimated blood loss (272.4 +/- 249.9 mL vs 174.8 +/- 140.8 mL; P <.001), hospital length of stay (10.15 +/- 4.85 days vs 7.05 +/- 5.14; P <.001), and overall complication rate (35.8% vs 22.4%; P =.015) were significantly less for revision procedures. Revision of more than one site on the graft resulted in longer operation time (P =.003) and estimated blood loss (P <.001), but similar complication rates (P = NS), compared with revision at only one site. Revisions that involved only the graft resulted in decreased hospital length of stay compared with revisions involving extension to native inflow or outflow vessels (P <.02). Return to preoperative ambulatory status at discharge was 71% after initial operation, and was 92% after revision (P <.001). Return to independent living at discharge was 66% after the initial operation, and was 80% after revision (P <.01). CONCLUSIONS: Operative revisions were better tolerated than initial vein graft procedures, but are still major procedures. Hospital length of stay is longer for patients undergoing proximal or distal extension of the graft to native vessels and in patients who are not ambulatory and living independently at discharge. Patients undergoing vein graft revision should be counseled about potential morbidity.

Long-term outcome of revised lower-extremity bypass grafts.

PURPOSE: Reversed lower-extremity vein grafts (LEVGs) frequently require operative revisions to maintain patency. Identifying grafts that are at risk, however, requires an intensive duplex scanning-based surveillance program. Excellent 5-year graft patency and limb-salvage rates have previously been reported in patients undergoing graft revisions, but results beyond 5 years are essentially unknown, a factor that is of importance in an increasingly aging population. This study was performed to determine the results of surgical revisions of LEVGs after a follow-up as long as 10 years. METHODS: All patients undergoing placement of a LEVG were observed in a program of duplex scanning-based surveillance as long as the patient remained a candidate for graft revision. Grafts were considered for revision on the basis of the presence of focal areas of increased velocity, a prestenotic to intrastenotic velocity ratio more than 3.0, or uniformly low velocities throughout the graft. All lesions were confirmed with preoperative arteriography before revision. Assisted primary patency, limb-salvage, and survival rates were determined by means of Kaplan-Meier analysis in all patients who underwent LEVG revision from January 1990 to December 2000. RESULTS: A total of 1498 LEVG procedures were performed during the study period. A total of 330 surgical graft revisions were performed on 259 extremities in 245 patients. The median follow-up period was 38 months. The assisted primary patency rate of all grafts, the limb-salvage rate for patients undergoing surgery for limb-salvage indications, and the survival rate of all patients were 87.4%, 88.7%, and 72.4%, respectively, 5 years after the original bypass grafting procedure, 85.7%, 83.4%, and 67.8%, respectively, 7 years after the original bypass grafting procedure, and 80.4%, 75.4%, and 53.4%, respectively, 10 years after the original bypass grafting procedure. A total of 180 revisions (55%) were performed during the first year, 110 (33%) between the first year and the fifth year, and 40 revisions (12%) were performed on grafts older than 5 years. LEVGs revised within the first year after bypass grafting had lesions within the graft in 78%, in the native arterial inflow in 10%, and in the native arterial outflow in 12%. This differed significantly from the location of lesions in revisions performed between 1 and 5 years and after 5 years (graft, 63% and 62%; inflow, 20% and 19%; outflow, 17% and 19%; P >.05, Chi-square). CONCLUSION: Excellent assisted primary patency and limb-salvage rates can be achieved for as long as 10 years in LEVGs that require revision, with only a 7% drop in overall patency and limb-salvage rates between the fifth and 10th years. Although most revisions were required within the first year, 34% were performed between the first year and the fifth year, and 11% after 5 years. These data support the growing body of evidence that favors an aggressive regimen of duplex scanning surveillance of LEVGs for the life of the graft. Revised grafts have excellent patency through 10 years.

Choice of autogenous conduit for lower extremity vein graft revisions.

BACKGROUND: Surgical revision to repair stenosis is necessary in about 20% of lower extremity vein grafts (LEVGs). Alternate conduit, especially arm vein, is often necessary to achieve a policy of all-autogenous revisions. Although basilic vein harvest necessitates deep exposure in proximity to major nerves, it typically uses a large vein unaffected by prior intravenous lines and as such appears ideally suited for revisions in which a segmental interposition conduit is needed for revision within the graft or for extension to a more proximal inflow or distal outflow site. In this report, we describe our experience with the use of the basilic vein for LEVG revisions compared with other sources of autogenous conduit. METHODS: All patients who underwent LEVG were placed in a duplex scan surveillance program. LEVGs that developed a focal area of increased velocity or uniformly low velocities throughout the graft with appropriate lesions confirmed with angiography were candidates for revision. All patients who underwent graft revision with basilic vein segments from January 1, 1990, to September 1, 2001, were identified, and their courses were reviewed for subsequent adverse events (further revision or occlusion) and complications of harvest. These revisions were compared with revisions in which cephalic and saphenous vein were used. RESULTS: One hundred thirty basilic veins were used to revise 122 LEVGs. The mean follow-up period after revision was 28 +/- 27 months. Ninety-three grafts (71%) remained patent with no further revision, and 37 grafts (29%) either needed additional revisions (22 grafts) or were occluded (15 grafts). Only four of these adverse events (11%) were directly attributed to the basilic vein segment. Ten of 43 grafts revised with cephalic vein (23%) were either revised or occluded, of which three were related to the cephalic vein segment (P = not significant, compared with basilic vein). Twenty-four of 81 grafts revised with saphenous vein (30%) were either revised or occluded, of which 11 were attributed to the saphenous vein segment (P <.01, compared with basilic vein). Two patients (1.5%) had complications from basilic vein harvest (one hematoma, one arterial injury). No neurologic injuries resulted from basilic vein harvest. CONCLUSION: The basilic vein is a reliable and durable conduit when used to segmentally revise LEVGs. Stenoses rarely occur within interposed basilic vein segments, and excellent freedom from subsequent revision or occlusion is possible. We conclude the basilic vein can be safely harvested with minimal complications and is ideally suited for use as a short segment interposition graft for LEVG revision.

Modifiable patient factors are associated with reverse vein graft occlusion in the era of duplex scan surveillance.

OBJECTIVE: Modifiable patient factors that contribute to graft occlusion may be addressed after surgery. To determine risk factors associated with reverse vein graft (RVG) occlusion, we examined the characteristics and duplex scan surveillance (DS) patterns of patients with RVGs. METHODS: Patients treated with RVG from January 1996 through December 2000 were identified from a prospective registry. The study population consisted of all patients with RVGs performed during the study period with grafts that subsequently occluded. Patients whose grafts remained patent served as age-matched and gender-matched control subjects. The prescribed DS regimen was every 3 months for the first postoperative year and every 6 months thereafter. Early DS failure was defined as having no DS within the first 3 months. Cox proportional hazards analysis was used to compare the two groups. Hazard ratios were calculated. RESULTS: During the study period, 674 patients underwent RVG. Fifty-five patients with occluded RVGs were compared with 118 with patent RVGs. The follow-up period for occluded grafts was 13.40 +/- 12.59 months and for patent grafts was 32.40 +/- 15.61 months. Dialysis therapy, a known hypercoagulable state, continued smoking, and DS failure were independent factors associated with RVG occlusion. The hazards ratio for dialysis was 6.45 (95% CI, 3.07 to 13.51; P <.001), for current smoking was 4.72 (95% CI, 2.5 to 8.85; P <.001), for hypercoagulable state was 2.99 (95% CI, 1.47 to 6.10; P =.003), and for early DS failure was 2.43 (95% CI, 1.29 to 4.59; P =.006). CONCLUSION: Continued smoking and failure to undergo DS within the first three postoperative months are modifiable factors associated with RVG occlusion. Smoking cessation and graft surveillance must be stressed to optimize patency of infrainguinal RVGs.

Postoperative duplex scan surveillance of axillofemoral bypass grafts.

OBJECTIVE: Duplex scan surveillance (DS) for axillofemoral bypass grafts (AxFBGs) has not been extensively studied. The intent of this study was twofold: 1, to characterize the flow velocities within AxFBGs; and 2, to determine whether postoperative DS is useful in assessment of future patency of AxFBGs. METHODS: We identified all patients who underwent AxFBG procedures between January 1996 and January 2001 at our combined university and Veterans Affairs hospital vascular surgical service. All grafts were performed with ringed 8-mm polytetrafluoroethylene with the distal limb of the axillofemoral component anastomosed to the hood of the femoral-femoral graft. DS was every 3 months for 1 year and every 6 months thereafter. Duplex scan results were compared in primarily patent grafts with grafts that thrombosed. Graft failures from infection were excluded. Influences of ankle-brachial index, blood pressure, outflow patency, operative indication, and comorbidities on graft patency were analyzed. RESULTS: One hundred twenty patients underwent AxFBG procedures. Twenty-eight were excluded because of infection or death before surveillance examination. Fourteen were lost to follow-up, 23 had failed grafts from occlusion, and 55 had grafts that remained patent. In the 78 patients evaluated during long-term follow-up period, the mean peak systolic velocities (PSVs) at the proximal (axillary) anastomosis during the first postoperative year ranged from 153 to 194 cm/s. Mean PSVs at the mid portion of the axillofemoral graft during the first postoperative year ranged from 100 to 125 cm/s, whereas those for the distal axillofemoral anastomosis ranged from 93 to 129 cm/s. Mean midgraft and distal anastomotic velocities obtained before thrombosis were significantly lower in the thrombosed grafts compared with the last recorded velocities at the same sites in the patent grafts (mean PSV, 84 versus 112 cm/s; P =.015; mean PSV, 89 versus 127 cm/s; P =.024, respectively). Forty-eight percent of occluded grafts had a mean midgraft PSV at last observation of less than 80 cm/s. Blood pressure correlated with midgraft velocity (r = 0.415; P <.05). With multivariate logistic regression analysis, a mean midgraft velocity less than 80 cm/s was the sole independent factor associated with graft failure (P <.01). No patients with midgraft velocities greater than 155 cm/s had occlusion. CONCLUSION: Flow velocity varies widely within and among AxFBGs. Patency of AxFBGs is associated with higher midgraft PSV, and thrombosis with midgraft velocities less than 80 cm/s.

Relationship of hemodialysis access to finger gangrene in patients with end-stage renal disease.

OBJECTIVE: We report a comprehensive review of our patients on hemodialysis with end-stage renal disease (ESRD) with finger gangrene to determine etiology, natural history, and prognosis of this condition. METHODS: Patients with ESRD with finger gangrene were identified from our computerized vascular registry. Presence of an ipsilateral arteriovenous fistula was determined, and patients were compared with a group of patients with ESRD without finger gangrene. Management consisted of arteriography, selective arteriovenous fistula management, and finger amputation. A multivariate analysis to determine risk factors associated with finger gangrene was performed. Repeat finger amputation and survival rates were determined with life-table analysis. RESULTS: Twenty-three patients (mean age at start of dialysis, 53 years) with finger gangrene were identified, with 48% (n = 11) having a functional ipsilateral arteriovenous fistula. Arteriography was consistent with diffuse atherosclerosis involving the radial, ulnar, palmar, and digital arteries precluding attempts at distal arterial bypass. Repeat finger amputations were necessitated in 52% of patients (n = 12), and bilateral finger gangrene developed in 61% of patients (n = 14). Starting dialysis at age less than 55 years (P =.0004), diabetes (P =.001), coronary artery disease (P =.0212), and lower extremity arterial occlusive disease (P <.0001) were significantly associated with finger gangrene. CONCLUSION: The young diabetic patient with diffuse vascular disease and ESRD is at high risk for the development of finger gangrene on chronic hemodialysis. Finger gangrene is the result of distal atherosclerosis and is not primarily related to dialysis access.

Relationship between site of initial symptoms and subsequent progression of disease in a prospective study of atherosclerosis progression in patients receiving long-term treatment for symptomatic peripheral arterial disease.

PURPOSE: There have been few studies of the natural history of peripheral arterial disease (PAD), and none have used serial noninvasive laboratory examinations for the objective quantification of disease progression. The relationship between the site of initial symptoms of PAD (lower-extremity disease [LED] vs cerebrovascular disease [CVD]) and the site of subsequent symptomatic progression (LED vs CVD vs coronary heart disease [CHD]) has not been examined. METHODS: This is a long-term, blinded prospective clinical research study of the relationship of PAD progression to multiple clinical, laboratory, and noninvasive vascular laboratory parameters. Patients with symptomatic LED, CVD, or both underwent comprehensive risk-factor assessment and were seen every 6 months for follow-up examinations. In addition to history and physical examination, all subjects underwent serial noninvasive lower-extremity and carotid artery testing. The relationship between the initial symptomatic site(s) and subsequent progression was examined by means of multivariate proportional hazards analysis, which was adjusted for age, diabetes mellitus, hypertension, smoking, cholesterol, homocysteine level, lowest initial ankle/brachial index (ABI), worst carotid stenosis, ABI progression, and carotid stenosis progression, because each of these factors was significantly associated with one or more aspects of progression. RESULTS: There were 397 study subjects (mean age, 66 years; 38% women) with a mean follow-up period of 48.5 months. LED was initially present in 88% of subjects and CVD in 37% of subjects (both were present in 25% of subjects). There were 78 deaths, 47 (60%) of which were caused by cardiovascular disease (18% mortality rate after 5 years by means of life table). Progression of disease as documented by means of vascular laboratory findings occurred in 90% of subjects by means of life table after 5 years (ABI progression, 31%; carotid stenosis progression, 40%). Symptomatic clinical progression of disease occurred in 52% of subjects by means of life table after 5 years (LED progression, 22%; CVD progression, 23%; CHD progression, 31%). By means of multivariate analysis, no significant relationship was found between the site of initial symptoms of PAD and the site(s) of subsequent symptomatic clinical progression (LED vs CVD vs CHD; P = not significant for all hazard ratios). CONCLUSION: Patients with symptomatic PAD experience symptoms of ongoing LED, CVD, and CHD with a frequency that is not influenced by the site(s) of their original symptoms. The hypothesis that lesions and resulting symptoms of systemic atherosclerosis occur at various anatomic sites as a matter of random chance should be tested with other studies.