Suggested standards for reports dealing with lower extremity ischemia

Periodicals Home

Search

User
Pref

Help

Logout

JVS Home

Table of
Contents

All
Issues

Order

About this
Journal

<<
Issue

>>
Issue

July 1986 • Volume 4 • Number 1

Special Article
Suggested standards for reports dealing with lower extremity ischemia

Prepared by the Ad Hoc Committee on Reporting Standards, Society for Vascular Surgery/North American Chapter, International Society for Cardiovascular Surgery. Robert B. Rutherford, M.D. (Chairman) [MEDLINE LOOKUP]
D. Preston Flanigan, M.D. [MEDLINE LOOKUP]
Sushil K. Gupta, M.D. [MEDLINE LOOKUP]
K. Wayne Johnston, M.D. [MEDLINE LOOKUP]
Allastair Karmody, M.D. [MEDLINE LOOKUP]
Anthony D. Whittemore, M.D. [MEDLINE LOOKUP]
J. Dennis Baker, M.D. [MEDLINE LOOKUP]
Calvin B. Ernst, M.D. [MEDLINE LOOKUP]
(members)
Nonmember consultants
Crawford Jamieson, M.D. [MEDLINE LOOKUP]
Shanti Mehta, M.S. [MEDLINE LOOKUP]

(W. L. Gore Company, Elkton, Md.).

Sections

Abstract
Definitions and classification criteria
Outcome criteria
Identifying and grading factors that modify outcome
Categorization of operations and procedures
Reporting deaths and complications
Final comment
References
Publishing and Reprint Information
Articles with References to this Article

• Previous article in Issue
• Next article in Issue
• Drug links from Mosby's DrugConsult
• Genetic information from OMIM

Abstract

TOP

Reports in the vascular surgery literature are often difficult to assess and compare with each other because of poorly defined terms, imprecise categorization, lack of indices for gauging the severity of the disease or the presence of risk factors capable of affecting outcome, and varying criteria for success or failure—in essence, a lack of standardized reporting practices. The joint councils of the Society for Vascular Surgery and the North American Chapter of the International Society for Cardiovascular Surgery have appointed an ad hoc committee to deal with this problem. This report represents the recommendations of the first of its several subcommittees, that is, the one dealing with reports on lower extremity ischemia. Certain terms are defined and criteria offered for uniformly gauging the severity of disease, the findings of diagnostic studies, the types of therapeutic interventions, and the outcome of such treatments. Although future modifications may further improve on this effort, it is hoped that this committee's recommendations will help establish reporting standards for articles dealing with lower extremity ischemia. (J VASC SURG 1986;4:80-94.)

Definitions and classification criteria

TOP

Although it is understood that progression of disease in a chronically ischemic extremity not infrequently takes place in a stepwise fashion, with each step representing an acute occlusive event, reports dealing with the management of lower extremity ischemia should not mix the management of these or other acute ischemic episodes with interval intervention for chronic ischemia because the results of emergency and elective interventional procedures are influenced by different variables and are not comparable. Thus, different classification criteria should be used for acute and chronic ischemia when attempting to stratify limbs according to severity of ischemia.

Acute (onset or progression of) ischemia

The following categories or gradations of severity of acute diffuse limb ischemia are recommended (Table I).

**Table I.** Clinical categories of acute limb ischemia
					Doppler signals
Category	Description	Capillary return	Muscle weakness	Sensory loss	Arterial	Venous
Viable	Not immediately threatened	Intact	None	None	Audible (AP >30 mm Hg)	Audible
Threatened	Salvageable if promptly treated	Intact, slow	Mild, partial	Mild, incomplete	Inaudible	Audible
Irreversible	Major tissue loss, amputation regardless of treatment	Absent (marbling)	Profound, paralysis (rigor)	Profound, anesthetic	Inaudible	Inaudible
AP = ankle pressure.

Viable: not immediately threatened; no ischemic pain, no neurologic deficit, skin capillary circulation adequate; clearly audible Doppler pulsatile flow signal in pedal arteries or ankle pressure above 30 mm Hg
Threatened viability: implies reversible ischemia and a limb salvageable without major amputation if arterial obstruction promptly relieved; ischemic pain and/or mild and incomplete neurologic deficit present (e.g., sensory loss involving only vibration, touch, position, or weakness of toe/foot dorsiflexion); “pulsatile flow” in pedal arteries not audible with Doppler instrument but venous patency demonstrable
Major, irreversible ischemic change: will require major amputation regardless of therapy; profound sensory loss and muscle paralysis, absent capillary skin flow or evidence of more advanced ischemia (e.g., muscle rigor or skin marbling); neither arterial nor venous flow signals audible distally

COMMENT: It is understood that some of these definitions and criteria are subjective and others arbitrary. Temporal criteria (e.g., 6 to 12 hours of ischemia) are not offered because tissue viability and damage also depend on location of occlusion and collateral circulation. More definitive tests of tissue viability are needed and hopefully will emerge. At this time, “reversibility” of ischemia or “salvageability” of the foot or limb cannot always be accurately predicted prospectively even with considerable clinical acumen. Nonetheless, grouping patients into viable, threatened, and irreversible categories is of value not only in determining appropriate therapy but in comparing the results of treatment.

It is further recommended that cases of arterial thrombosis and embolism not be indiscriminately mixed together or, if they are combined, the distribution of cases into these categories be included for comparative purposes. Cases of atherothrombotic microembolism (“blue toe syndrome”) usually present with transient focal ischemia with occasional minor tissue loss but without diffuse forefoot ischemia and should be excluded entirely or included in category 1. As discussed later in more detail, the practice of including such cases of transient focal ischemia in either the “threatened” or “limb salvage” categories is condemned.

Chronic ischemia

Before offering classification criteria, a number of terms deserve definition and clarification, judging by current literature usage.

“Claudication” implies extremity pain, discomfort, or weakness consistently produced by the same amount of walking or equivalent muscular activity in a given patient and is promptly relieved by cessation of that activity. Ordinarily claudication implies ischemic muscle pain induced by exercise and as such may be identified as hip, buttock, thigh, or calf claudication. However, there is also a form of foot claudication in which pedal ischemia induced by exercise causes not only pain but numbness. In these patients, the occlusive lesions are more distally located, and the degree of ischemia is more severe.

The severity of claudication can be reliably related to time or distance walked only if speed and incline grade are also standardized. Even walking distance is subject to sufficient variation that its use to judge the therapeutic effect of pharmacologic treatment has been challenged. The current practice of categorizing operative candidates as “disabled” or “less than one block” claudicators is imprecise. Disability is relative, being related to activity levels governed by age, occupation, and avocational interests. Therefore, “disabling claudication,” although an acceptable indication for operation in carefully selected patients, is no more acceptable as a categorizing criterion than “less than one block” claudication. Either the broad category of intermittent claudication should be used or objective and reproducible criteria for further break-down should be established.

Claudicators need further separation only for the comparative purposes of clinical investigation. Resting ankle systolic pressure measurements will not cleanly separate claudicators according to degree of severity. The suggested noninvasive vascular laboratory test criteria (Table II) were arbitrarily chosen to represent first the minimum acceptable objective evidence of true claudication, designated as “mild”; then “moderate” and “severe” were separated by whether or not the patient can complete 5 minutes on the treadmill and whether or not the reduction in ankle pressure occurring after this exercise (or an equivalent degree of induced hyperemia) reaches a level commonly associated with disabling claudication, that is, 50 mm Hg.

**Table II.** Clinical categories of chronic limb ischemia

AP = ankle pressure; BP = brachial pressure; PVR = pulse volume recording; TP = toe pressure; TM = transmetatarsal. *Five minutes at 2 mph on a 12% incline.

Claudication may be experienced without the ankle pressure being reduced to this level but it is usually not very disabling. For classification as a severe claudicator, lesser speeds or duration of ambulation are acceptable if typical claudication discomfort forces cessation of exercise within 5 minutes and the ankle pressure drops below 50 mm Hg.

Wilbur and Olcott¹ have shown, in testing claudicators, that the ankle pressures obtained 1 minute after a 5-minute treadmill exercise are roughly equivalent to those obtained 30 seconds after the lesser degree of hyperemia induced by an equal duration of suprasystolic thigh cuff occlusion. Repeated dorsiflexion of an elevated limb was significantly less effective in producing a pressure drop. Therefore, the former two are acceptable as equivalent stress tests. The speed of 2 mph, with an incline of 12%, is used commonly enough to be accepted as standard for treadmill testing. Five minutes on an inclined treadmill at 2 mph is almost equivalent to walking two blocks at a moderate speed.

“Ischemia rest pain” or diffuse pedal ischemia has been nicely characterized by Cranley.² It is a severe pain not readily controlled by analgesics and localized, in the chronically ischemic extremity, in the forefoot and toes, or, if more proximal, at least does not spare these distal parts. It may also be localized to the vicinity of focal ischemic lesions. It may be brought on or made worse by elevation and relieved by dependency and therefore, is often only experienced at night or when reclining. Diffuse pedal ischemia is commonly associated with ankle pressures below 40 mm Hg and toe pressures below 30 mm Hg.

“Gangrene” may be focal, as in the case of focal thrombosis or atherothrombotic microembolism, so that there is still adequate perfusion of adjacent tissues to allow successful auto- or surgical amputation. Such focal gangrene is often not associated with diffuse pedal ischemia and typical rest pain. Gangrene associated with diffuse pedal ischemia will not allow successful management by local measures and will invariably be associated with typical ischemic rest pain.

Ulcers in distal parts of the extremity may be caused, or made to persist, by one or more etiologic factors, each with its own distinguishing characteristics (e.g., pressure, venous insufficiency, trauma, diabetic or other neuropathies), as well as by persistent arterial insufficiency. The term “nonhealing ischemic ulcer” implies that, regardless of origin, there is insufficient arterial perfusion to support the inflammatory response required for healing. Associated with this, there should be ischemic rest pain and objective evidence of diffuse forefoot ischemia (e.g., critical reductions in the ankle or toe pressures, a flat or barely pulsatile plethysmographic tracing at the ankle or transmetatarsal level, or the lack of an inflammatory response as gauged by radionuclide studies).^3-6 An ankle pressure upper limit of 60 mm Hg is offered in this category rather than 40 mm Hg. Similarly, a toe pressure of 40 mm Hg is suggested for this category instead of the 30 mm Hg suggested for rest pain, in recognition of the additional perfusion (i.e., inflammatory response) required to heal an ulcer or a distal amputation, especially if infection is present.^3,6

“Limb salvage” is a regrettable term in that it is a misnomer and is often loosely applied; it would best be abandoned. Salvage of the foot, not the limb, is the ultimate criterion and this is retrospectively determined, yet the term is often applied prospectively. Chronic critical ischemia, as defined by Jamieson et al.,³ might be more apt but common usage may dictate persistence of the former term for the present. It would even be better to use the term “foot salvage,” for what is implied is the otherwise inevitable loss of the foot and that a major amputation has been (or can be) averted only by successful arterial reconstruction or other therapeutic intervention. The presence of signs or symptoms of critical ischemia (e.g., rest pain, nonhealing ulceration, or gangrene) plus objective evidence of diffuse pedal ischemia as defined earlier (Table II) qualify the patient for such categorization. However, for rest pain in the absence of frank tissue loss, at least 6 weeks should be allowed for the development of collateral circulation before such designation.

In this context, if the outcome involves minor amputation when major amputation would otherwise have been inevitable, it may be included under “foot salvage.” In this regard, minor and major amputation should be defined. “Minor amputation” implies retention of a functional foot, one which would allow standing and walking without a prosthesis. A modified shoe is allowable but a Syme's amputation involves shortening and prosthetic fitting and thus would not qualify as a minor amputation and inclusion under “foot salvage.” Therefore, for the most part, minor amputation will be represented by digital or transmetatarsal amputations. Syme's amputations should be included under “major amputations.” Revascularization that allows healing of a below-knee amputation when above-knee amputation would have been otherwise predicted, although in a sense representing partial limb salvage, does not qualify under the designation “foot salvage.”

Operations for microembolism or “blue toe syndrome.” although often justified to save the foot from eventual partial or complete loss after recurrent embolization, do not qualify for inclusion with “foot salvage” operations unless there is objective evidence of diffuse pedal ischemia, a visible threat of tissue loss, and a proximal hemodynamically significant obstructive lesion is corrected or bypassed. Because of their uniqueness, such cases are better reported separately. If included in overall reviews of experiences with arterial reconstructions, those without diffuse pedal ischemia should be listed with other hemodynamically insignificant lesions (grade or category 0), like certain graft structural defects, unless of course they are actually associated with an occlusive lesion causing claudication. Finally, it would seem advisable to separate or indicate the relative proportion of nonhealing ulcers and gangrene in those with actual tissue loss.

A suggested classification for grading the severity of chronic arterial occlusive disease for the purposes of standardized reporting practices is outlined in Table II with these definitions and criteria. Symptomatic disease has been subdivided into six categories to provide the greater definition required for more discriminating research but simpler broad gradations, on the basis of Fontaine's original clinical staging, are also recognized. A zero category or grade has been added to include those with no symptoms, mere cold sensitivity, no clinical evidence of disease, or lesions of no hemodynamic significance. Such a category is valuable because it also allows postoperative improvement to be gauged.

Outcome criteria

TOP

Criteria for significant change in status (improvement, deterioration, or failure)

Although patency is accepted as the ultimate criterion of success when results of arterial reconstruction are reported, this is primarily because it is a discrete and comparable end point. However, situations exist in which patency does not necessarily mean success (e.g., an aortofemoral bypass performed in the face of such significant downstream occlusive disease that the patient is not relieved of claudication, rest pain, or the need for major amputation, although the graft is clearly patent). This may be termed a “hemodynamic failure.” At the other extreme is the not uncommon situation, judging from the literature, in which a bypass graft performed for “limb salvage” occludes but the limb is no longer threatened. How often such cases are the result of loosely defined indications and how often they represent true salvage because of time gained for collateral development or sufficient improvement in circulation to allow healing remains grounds for debate. Suffice it to say that clear definitions of patency, foot salvage, and/or “significant” improvement are needed to serve as ultimate measures of success, and reporting the rate of other criteria in addition to patency will provide greater perspective and better grounds for comparison.

There is significant advantage, in comparing certain procedures, to gauging the relative degree of clinical improvement. For reporting purposes, the designation “significantly improved” ordinarily requires upward shift by at least one clinical category (Table II ) but those with actual tissue loss (category 5) must move up at least two categories and reach the level of claudication to be considered improved. In addition, some objective criteria of improvement should be included (i.e., a change in the ankle/brachial index [ABI] of more than 0.10). Comparing mean elevations of the ABI alone may not be very discriminating. Therefore, the following scale has been suggested for gauging degree of improvement or worsening. Its use may provide group indices for comparison.

+3 = markedly improved: symptoms gone or markedly improved; ABI increased to more than 0.90
+2 = moderately improved: still symptomatic but at least single category*
*Categories refer to clinical classification (Table II ). improvement; ABI increased by more than 0.10 but not normalized
+1 = minimally improved: greater than 0.10 increase in ABI but no categorical improvement, or vice versa (i.e., upward categorical shift without an increase in ABI of more than 0.10)
0 = no change: no categorical shift and less than 0.10 change in ABI
–1 = mildly worse: no categorical shift but ABI decreased more than 0.10, or downward categorical shift with ABI decrease less than 0.10
–2 = moderately worse: one category worse or unexpected minor amputation
–3 = markedly worse: more than one category worse or unexpected major amputation

Criteria for patency

Articles in scientific journals should not accept patency rates that are not made on the basis of objective findings. “No evidence of occlusion” cannot be equated with patency for reporting purposes. A bypass graft or otherwise reconstructed arterial segment may be considered patent when any one of the following five criteria is met. Beyond the last date of such proof of patency, they must be considered lost to follow-up.

Demonstrably patent by conventional arteriography or some other established imaging technique (e.g., digital subtraction arteriography, ultrasound, radionuclide study, or magnetic resonance imaging)
Maintenance of the achieved improvement in the appropriate segmental limb pressure index, which, if not normalized, must be at least 0.10 above the preoperative index and no more than 0.10 less than the maximum postoperative index; the former without the latter qualifies as “deterioration” rather than “failure.”
Maintenance of a plethysmographic tracing or oscillometric reading distal to the reconstruction, which is significantly greater in magnitude than the preoperative value (This is acceptable only when accurate pressures cannot be measured, as with calcific arteritis in a diabetic patient. For pulse volume recorders, this is +5 mm or +50%; for an oscillometer + 1/2 unit.)
The presence of a palpable pulse, or the recording of a biphasic or triphasic Doppler wave form at two points directly over a superficially placed graft.
Direct observation of patency at operation or postmortem examination

COMMENT: Although palpable pedal pulses readily felt by an experienced observer are clearly adequate for routine clinical assessment, comments to this effect made in the patient's record by nurses, residents, or fellows cannot be extracted as proof of patency for reports in scientific journals. Palpation of pulses by two observers, or palpation of a pedal pulse synchronous with the heart rate monitored by a second observer are clearly also clinically acceptable means of avoiding false pulse detection but are not suitable for uniform application. Accurate patency data are so crucial to comparisons of arterial reconstructive techniques that discriminating methods deserve to be used. Doppler measurement of ankle and brachial pressures takes only a few minutes.

Patency status: primary vs. secondary patency

With the help of graft thrombectomy or thrombolysis, revision or “redo,” it may be claimed that the original graft is still patent. It is important in this regard to separate “primary” from “secondary” patency. The graft is considered to have “primary” patency if it has had uninterrupted patency with either no procedure performed on it or a procedure, such as transluminal dilation or proximal or distal extension to the graft, to deal with disease progression in the adjacent native vessel. Thus, the only exceptions that do not disqualify the graft for primary patency are procedures performed for disease beyond the graft and its two anastomoses. Dilations or minor revisions performed for stenoses, dilations, or other structural defects before occlusion do not constitute exceptions as they are intended to prevent eventual graft failure.

If graft patency is restored after occlusion by thrombectomy, thrombolysis or transluminal angioplasty, and/or problems with the graft itself or one of its anastomoses require revision or reconstruction, this must be listed under “secondary” patency. A “redo,” as defined later, does not contribute to secondary patency, since most of the original graft is not retained in continuity.

It should be understood that both primary and secondary rates are important. The former is important in judging the natural history of a graft or reconstructive procedure, and the latter is important to indicate the long-term function that can be achieved with the aid of secondary or adjunctive procedures. Both provide valuable information but when only one or the other patency rate is presented, and which one not clearly identified, comparison between different reports on the same type of reconstructive procedure is difficult. Therefore, it is recommended that in each report, both primary and secondary patency rates be presented and the patency rate under discussion is qualified as primary or secondary.

Estimating patency rates

Although subject to some artifact, so that projected and actual patency rates (e.g., patent at the end of 5 years)⁷ are not necessarily the same, the life-table method is probably the best commonly used way of presenting patency data on patients who are coming for operation at different points in time and are followed for different time intervals. The life-table method has been clearly explained by an international committee of 10 eminent biostatisticians in two companion articles published in the British Journal of Cancer.^8,9 It uses and abuse in vascular surgery have been thoroughly unmasked.¹⁰

General guidelines on life-table analysis

One may use any standard text¹¹ or the references cited earlier to do the life-table computations. The reporting interval for patients lost to follow-up or those who died with parent grafts stops at the time of their last examination. Patients whose grafts have failed since their last examination are statistically treated as having failure dated halfway between two examinations. Thus, if the frequency of examination is increased, the accuracy of the data is improved.

Life-table analysis should include the following columns in the table: intervals in months, number at risk at the start of the period, number failed during the period, number withdrawn patent because of death or being lost to follow-up, interval patency, cumulative patency, and standard error (see Table III and Fig. 1 for working example).

**Table III.** A typical life-table analysis
			No. withdrawn patent due to
Interval (mo)	No. of grafts at risk at start	No. of failed grafts	Duration	Loss to follow-up	Deaths	Interval patency rate	Cumulative patency (%)	Standard error (%)
0-1	57	10	0	0	0	0.82	100	0
1-3	47	3	3	0	1	0.93	82	5.1
3-6	40	4	2	0	0	0.90	76	5.9
6-9	34	4	2	1	1	0.88	69	6.6
9-12	26	0	3	0	2	1.00	60	7.4
12-15	21	2	1	1	1	0.90	60	8.3
15-18	16	0	2	0	0	1.00	54	9.2
18-21	14	1	2	0	0	0.92	54	9.8
21-24	11	2	2	0	0	0.80	50	10.7
24-27	7	1	1	0	0	0.85	40	11.7
27-30	5	0	0	0	1	1.00	34	12.4
30-33	4	0	1	0	0	1.00	34	13.8
33-36	3	0	1	0	0	1.00	34	15.9
NOTE: Type of operation, tibial bypass; No. of patients, 50; No. of grafts, 57. For explanation of columns, see text.

Fig. 1. Acceptable ways of graphically presenting the life-table data recorded in Table III. A, Numbers at risk at beginning of each period are indicated. Cumulative patency rate is indicated by stepped rather than curved line, and line is interrupted beyond point at which standard error exceeds 10%. B, Standard errors are also graphically shown. This method of display may be preferable if two patency curves are being compared in the same figure.

Click on Image to view full size

Cumulative mortality, although not a requisite part of the life table, adds valuable perspective and deserves inclusion.

The following paragraphs explain the determination of the different columns in the life table used in this article (Table III).

Interval in months

Intervals can be chosen to represent any desired time span equal to or shorter than the review period and they need not be equal. However, it is important to ensure that intervals do not overlap. For example, in Table III there were 10 of a total of 57 grafts whose follow-up times since surgery were less than but not including one month. It is always useful to have the first interval as 0 to 1 month to show early patency. Thereafter, three monthly or six monthly intervals may be chosen. More frequent reviews help define the life-table more precisely.

No. of grafts at risk at start

For the first interval this shows the number of grafts in the study. For the second interval, the number at risk at start is obtained by subtracting the sum of columns 3, 4, 5, and 6 in the first interval from the corresponding figure in column 1.*

*Columns are considered as numbered 1 through 9, from left to right. The procedure continues and each successive number in column 2 is obtained similarly.

No. of failed grafts

This column shows the number of grafts that closed having their times to failure falling in the respective intervals.

Duration

This column gives the number of patent grafts having their follow-up times in the respective intervals.

Loss to follow-up

These numbers represent patients whose whereabouts are not know and whose last examination revealed patent grafts with follow-up times falling in the intervals shown.

Death

These numbers indicate the patients who have died with patent grafts and at the last examination their follow-up times were as shown in the table.

Interval patency rate

The numbers in this column for interval patency need some explanation. In calculating interval patencies, an assumption is made that the patients withdrawn patent (for death, loss to follow-up, or because of limited time elapsed since operation) were all withdrawn precisely at the midpoint of the interval range. It is also assumed that during the half of the interval for which they were not followed up, they were subject to half the risk of failure of the entire interval. On these assumptions, it can be shown that:
Interval failure rate = number failed ÷ (number at risk –½ the total number withdrawn)
Interval failure rate = column 3 ÷ (column 2 –½ the sum of columns 4, 5, 6) and the interval patency = 1 – interval failure rate*

*Columns are considered as numbered 1 through 9, from left to right.

Cumulative patency

The percentage of cumulative patency represents the proportion of grafts remaining patent at the start of each interval. For example, immediately after surgery, the cumulative patency is 100% for the first interval. For each successive interval, this figure is obtained by multiplying the interval patency rate by the percentage of cumulative patency in the preceding interval. For example, as shown in Table III , the cumulative patency for the second interval is 0.82 multiplied by 100 or 82%. This means that 82% of the grafts would remain patent for 1 month or more.

Standard error

The approximate standard errors are computed using the simple formula recommended by Peto et al.^8,9 According to this formula:

where L is the cumulative patency expressed as a fraction and N is the number still at risk at the start of the interval. When standard errors are shown on the life-table graph to show uncertainty of estimates of cumulative patencies at different points in time, plus or minus two times standard error should be used to show an approximate 95% confidence range.

Complete life-table data should be submitted in table form to allow analysis, even if the patency rates are also graphically illustrated. The graphs may be published without the tables at the discretion of journal editors. When patencies determined by life-table analysis are presented graphically, either numbers at risk at the start of each interval must be shown or the standard error of each estimate of patency must be displayed, preferably both. This allows the reliability of the reported patencies to be judged. For joining cumulative patency points on the graph, the procedure recommended in the British Journal of Cancer citation should be used. This requires that points should be joined in steps and not by a continuous line.

When the standard error of the patency estimate exceeds 10%, the curve beyond that point should be omitted or at least represented by a dotted line or in some other manner to indicate poor reliability of estimates beyond that point. All reconstructive failures, including early thrombosis or removal for sepsis, although they may have been due to technical or other errors, must be included in the computation of the life-table analysis of primary patency. Both primary and secondary patencies can be represented on the same graph. All comparisons of life-table estimates must be done with the log rank test of significance. It is not correct to compare patencies at specific intervals with the standard errors. The simple log rank test compares two or more life tables over the entire period of observation.

Whenever possible, separate life tables should be provided for each type of operative procedure. Generally, one should not mix several different infrainguinal bypass procedures but report separately above-knee femoropopliteal bypass, below-knee femoropopliteal bypass, and femoral infrapopliteal bypass, as examples. Where pertinent (i.e., where differences are claimed), additional life-table analyses should be reported for different indications for operation (claudication vs. foot salvage), for different runoff conditions or for any major risk or treatment factors that affect patency (e.g., with and without antiplatelet therapy or diabetic vs. nondiabetic). In some instances, interdependence of variables will limit the confidence with which conclusions may be drawn from this subgrouping of data.

Identifying and grading factors that modify outcome

TOP

Clinical reports evaluating revascularization procedures, particularly those comparing different treatment modalities, may be difficult to interpret when differences in factors that can affect outcome are not identified and characterized. For example, diabetes, tobacco usage, and occlusive disease distal to the revascularization (“runoff”) may affect patency rates and degree of improvement, whereas cardiac, pulmonary, and renal status may influence operative mortality and long-term survival. Grading such factors in severity, with definitions for (1) mild, (2) moderate, and (3) severe, would provide severity indices for intergroup comparison. The following simplified grading system is offered for common risk factors, recognizing that many alternative schemes have been proposed but none universally accepted. A scheme for grading “runoff” is also proposed.

Diabetes: 0 = none; 1 = adult onset, diet-controlled; 2 = adult onset, insulin-controlled; 3 = juvenile onset.

Tobacco use*

*0 = absent, none, negligible; 1 = mild; 2 = moderate; or 3 = severe.: 0 = none or none for last 10 years; 1 = none current, but smoked in last 10 years; 2 = current,†

†Includes abstinence less than 1 year. less than 1 pack/day; 3 = current, greater than 1 pack/day.

Hypertension: 0 = none‡

‡Cutoff point, diastolic pressure regularly above or below 90 mm Hg. 1 = easily controlled§

§Determined by noninvasive test or arteriogram. with single drug; 2 = controlled with two drugs; 3 = requires more than two drugs or uncontrolled.

Hyperlipemia: 0 = cholesterol/triglycerides within normal limits for age; 1 = mild elevation, controllable by diet; 2 = types II, III, or IV, requiring strict dietary control; 3 = requiring dietary and drug control.

Cardiac status: 0 = asymptomatic, normal electrocardiogram (ECG); 1 = asymptomatic, remote myocardial infarction (MI) by history (greater than 6 months), occult MI by ECG; 2 = stable angina, controlled ectopy or asymptomatic arryhthmia, drugcompensated congestive heart failure (CHF); 3 = unstable angina, symptomatic or poorly controlled ectopy/arrhythmia, poorly compensated CHF, MI within 6 months.

Carotid disease: 0 = no symptoms, no bruit, no evidence of disease; 1 = asymptomatic but with evidence of disease§; 2 = transient or temporary stroke; 3 = completed stroke with permanent neurologic deficit.

Renal status*

*Refers to stable levels, not transient drops from hydration or response to arteriography.: 0 = no known renal disease, serum creatinine level less than 1.5 mg/dl, creatinine clearance greater than 50 ml/min; 1 = creatinine, 1.5 to 3.0 mg/dl, creatinine clearance 30 to 50 ml/min; 2 = creatinine, 3.0 to 6.0 mg/dl, creatinine clearance, 15 to 30 ml/min; 3 = creatinine greater than 6.0 ml/dl, creatinine clearance less than 15 ml/min or on dialysis or with transplant.

Pulmonary status: 0 = asymptomatic, normal chest x-ray film, pulmonary function tests (PFTs) 20% of predicted; 1 = asymptomatic or mild dyspnea on exertion, mild x-ray parenchymal changes, PFTs 65% to 80% of predicted; 2 = between 1 and 3; 3 = vital capacity less than 1.85 L, FEV₁ less than 1.2 L or less than 35% of predicted, maximal voluntary ventilation less than 28 L/min or less than 50% of predicted, PCO₂ greater than 45 mm Hg, supplemental oxygen use necessary or pulmonary hypertension.

COMMENT: Goldman's Cardiac Risk Index combines nine independent factors that correlate with life-threatening and fatal complications.¹² Cardiac, pulmonary, and renal factors are included. In the absence of agreement on separate cardiac, pulmonary, and renal factors, this approach may offer an acceptable alternative. It is understood that it may not be appropriate or feasible to include all these risk factors in each report, but those claiming or stressing improved patency or mortality rates should support their claim by including such information on appropriate risk factors.

Runoff

It is understood that no scheme for grading runoff is perfect, likely to be universally accepted, or will always correlate with early or late graft failure. Nevertheless, a grading scheme that provides some degree of correlation with outcome is desirable and simple (good/poor or 1 to 4) grading schemes have proved inadequate. The one proposed here may be applied to all levels of distal anastomosis rather than just to femoropopliteal bypass. As seen in Table IV, it grades both the degree of occlusion and the relative contribution to outflow of each vessel from 0 to 3 and combines the two in a decimal system that assigns 1 to a widely patent runoff and 10 to an isolated, blind segment. In this scheme, higher values correspond to higher resistances so that resistances in series and in parallel (e.g., axillobifemoral and sequential bypasses) can be consistently graded. Calculations for some of the more complex examples of this scheme are illustrated in Figs. 2 and 3.

Fig. 2.A, Bolder numbers indicate value, out of possible total of 3, assigned to each vessel; numbers in finer print indicate values assigned to varying degrees of occlusion (Table IV). B, Runoff resistance values are calculated as follows for grafts entering the three levels indicated: Level A, superficial femoral = 2 × 3; profunda femoris = 1 × 0; total = 6 + 0 + 1 = 7. Level B, distal popliteal blind segment = 3 × 3 + 1 = 10. Level C, anterior tibial = 2 × 0; arch = 1 × 0; total = 0 + 0 + 1 = 1.

Click on Image to view full size

Fig. 3. In each of three graft configurations (A-C), individual runoff values for right and left limbs are 7 and 4, respectively. Values for aortobifemoral stem (A), entire right axillofemoral graft (B), and proximal axillofemoral stem (C) are all 2.5 (1/R = 1/4 + 1/7 = 11/28 = 1/2.5). However, distal axillofemoral limb (C) has runoff resistance value of 7.D shows a sequential bypass where resistance values of proximal and distal limbs are 10 and 1, respectively. Combined they give a resistance value for proximal stem of 0.9 (1/R = 10/10 + 1/10 = 11/10 = 1/0.9).

Click on Image to view full size

**Table IV.** Weighting of runoff arteries (total of three units)
	No. of units assigned*
Site of distal anastomosis (artery)	3	2	1
Common iliac		External iliac	Hypogastric
External iliac	Common femoral	SFA	Profunda femoris
Common femoral		SFA	Profunda femoris
Popliteal above-knee	Distal popliteal		Anterior tibial
			Posterior tibial
			Peroneal
Anterior tibial		Distal tibial	Pedal arch
Posterior tibial		Distal tibial	Pedal arch
Peroneal		Pedal arch	Collaterals to anterior and posterior tibial arteries
Dorsalis pedis	Pedal arch
SFA = superficial femoral artery. *Points assigned for degree of occlusion: 3 = occluded throughout most of its length; 2.5 = occluded for less than half its length with visible collaterals; 2 = 50% to 99% stenosed; 1 = 20% to 49% stenosed; 0 = normal or minimal evidence of disease (i.e., less than 20% narrowing).

Most calculations obviously will be much simpler than these.

COMMENTS:

Three weighting units are assigned to the total number of runoff arteries. The dominant of two runoff vessels is assigned two of the three units (e.g., superficial femoral = 2, profunda femoris = 1) whereas three more or less equal runoff vessels, like the three infrapopliteal arteries, are assigned one unit each (Table IV and Fig. 2, A).
With a maximum resistance value of three for each of the three units, a total of nine points is possible. This “resistance” estimate is usually added to a base “resistance” of one, in recognition of the fact that even a widely patent distal bed offers some resistance and also to avoid zeros in arithmetic calculations. A blind segment thus becomes a 10.
In multiple outflow bypasses, as in the two limbs of aortobifemoral and axillobifemoral bypasses, the reciprocal sum of each outflow artery is added in grading resistance in the proximal limb or stem, but each distal limb of the graft is graded alone for correlation with graft limb patency (Fig. 2).
In grading the pedal arch, zero is assigned for completely patent arch, 1.5 for partial occlusion, and 3 for little or no arch visualized.

Categorization of operations and procedures

TOP

Definitions

It is important to identify, if not separate, primary and secondary operations, principal and adjunctive procedures, and different types of procedures (e.g., reconstructive, restorative, nonreconstructive, and ablative). The following definitions are suggested and should be followed for uniform reporting.

A “primary operation” is the first operation of a given type ever performed on a particular arterial segment. Subsequent such operations performed on the same arterial segment are called “secondary operations.” For example, if a profundaplasty fails to save an ischemic foot and a femoropopliteal bypass is subsequently performed, both may be considered primary operations. However, if a femoropopliteal bypass is redone or extended down to a tibial artery, these would be called secondary operations. Such operations performed as the first operation at a different institution or by a different surgeon are still secondary operations.

A “principal procedure” is the one that the surgeon believes to be the most responsible or important in improving arterial circulation. An “adjunctive procedure” is any other simultaneous or subsequently planned procedure that is designed to augment the effects of the principal procedure, such as arterial dilation or profundaplasty. An “ancillary procedure” does not contribute to the overall effect, such as intraoperative arteriography.

An “elective operation” is one that is performed without urgency and scheduled during regular operating erating time at the mutual convenience of both the patient and the surgeon. An “urgent operation” is one that should be performed as soon as the minimum necessary preoperative preparation and diagnostic studies are completed. An “emergency operation” is one that must be performed as soon as possible (within a relatively short time, e.g., 4 hours) because of an immediate threat to limb or life.

A “reconstructive procedure” is one that is performed to remove an obstructive or aneurysmal lesion involving the arterial wall and/or to restore pulsatile flow beyond the involved arterial segment. This category would include bypass graft, interposition graft, resection and anastomosis, endarterectomy, or surgical angioplasty with and without patch graft.

A “restorative procedure” is one in which obstruction is removed from the arterial lumen of an otherwise normal arterial segment or when the lumen can be restored to near normal without direct reconstruction. Examples include thrombectomy, embolectomy, enzymatic thrombolysis, laser plaque destruction, and transluminal dilation.

A “nonreconstructive procedure” is any procedure designed to improve or to protect blood flow without direct arterial reconstruction. Included would be sympathectomy, fasciotomy, and release of compression of the artery as by division of a band or resection of a rib. “Ablative procedures” are those designed to remove nonviable or diseased material (tissue or graft) or that interrupt flow in patent vessels. These would include major or minor amputation, debridement or removal of an infected graft. Ligation of an arterial segment is included here although nothing is removed.

Finally, when the procedure fails and corrective measures are taken to restore functional patency, it is important to distinguish between revisions and “redo” procedures. In a “revision,” there is retention without significant modification of all or most of the graft or reconstructed segment. Distal extension from a graft that is or has been rendered patent throughout most of its length may be considered a revision rather than a “redo.” A “redo” or secondary reconstruction implies replacement or bypass of all or most of the graft or reconstructed segment.

Grouping and characterization of lower extremity revascularization procedures

In addition to the broad categories defined above, lower extremity vascularization procedures can be characterized by type of procedure (e.g., endarterectomy, interposition graft, bypass, or embolectomy), location with identification of the arterial segment operated on or the proximal and distal sites of a bypass or endarterectomy, and side (e.g., right, left to right, or bilateral). Additional specific details, such as graft type, shape, size, type of anastomosis, anatomic route of a bypass, and incisional approach, are worth documenting but may create so many variables that innumerable subgroups would be required and defy categorization. Therefore, there is a need for a more inclusive system of grouping, allowing similar procedures to be compared. The following general groupings are suggested for lower extremity revascularization procedures. This system divides procedures according to whether they are direct or indirect (extra-anatomic or ex situ) and whether they deal with occlusive disease proximal to, at, or distal to the femoral bifurcation.

Direct (in situ) proximal revascularization—includes aortoiliac, iliofemoral and aortofemoral endarterectomy or bypass, unilateral or bilateral
Indirect (ex situ) proximal revascularization—includes such extra-anatomic bypasses as axillofemoral, crossover femorofemoral, axillobifemoral, and thoracoaortofemoral
Direct (in situ) femoral revascularization—from external iliac to proximal superficial and profunda femoral arteries; includes profundaplasty by endarterectomy and/or patch angioplasty
Indirect (ex situ) femoral revascularization—includes obturator bypass, axillopopliteal, and crossover femoropopliteal bypasses
Distal revascularizations—includes above-knee (AK) femoropopliteal bypass, below-knee (BK) femoropopliteal bypass, femorocrural and femoropedal bypasses, AK and BK popliteocrural and AK and BK popliteopedal bypass and sequential femorodistal bypasses to popliteal, crural, or pedal arteries. Differentiation between an in situ or ex situ graft course here does not warrant subcategorization.

COMMENT: When comparing procedures that are “competitive” for revascularizing a particular arterial segment, as in comparing aortobifemoral and axillobifemoral bypass or surveying the overall results of surgical management of aortoiliac occlusive disease, it is appropriate that such different procedures be included in one report. On the other hand, when comparing two or more technical or treatment variables, such as graft types or antithrombotic drugs, with graft occlusion as the end point, it is recommended that broad categories containing multiple operations not be handled together. Instead, these variables should be compared on the same or similar procedures.

The category “femorofemoral bypass” should include all similar crossover grafts whether the origin be the external iliac artery or the termination the profunda femoris artery. Similarly, femoropopliteal and femorocrural grafts should include those that also originate on the external iliac, superficial femoral, or profunda femoris arteries. However, the proportion of each different origin and/or termination can and often should be noted. In the important instance of AK vs. BK popliteal termination of a graft, separation into two groups is preferable to simply noting the proportion of each and then evaluating them together. Sequential distal bypass grafts are better considered separately and not included with either femoropopliteal or femorocrural bypass (some reports have included each limb in each of these two categories). Patencies of multiple termination grafts (e.g., aortobifemoral, axillobifemoral, AK femoralpopliteal-tibial) should be calculated by considering each graft limb separately.

Reporting deaths and complications

TOP

Postoperative deaths may be due to physician errors (in diagnosis, technique, judgment, or management) or, in their absence, to disease of the patient. Late deaths are usually attributed to either the underlying disease, delayed complications of surgical management, or are considered “unrelated.” Both early (less than 30 days) and late (more than 30 days) mortality occurring after lower extremity revascularization procedures should be reported to give a truer perspective, and the additional breakdown offered earlier for late deaths is recommended.

Complications of lower extremity procedures can be either specific or nonspecific but the separation between the two is often indistinct. The nonspecific category includes such problems as atelectasis, dehiscence, and congestive heart failure as well as some that, although not specifically related to operative technique, are nevertheless indirectly related to the procedure or to the underlying disease it treats. Examples are myocardial infarction, stroke, deep venous thrombosis, and/or pulmonary embolism. Even such universal complications as wound infection and hemorrhage may be relatable to specific aspects of the patient's disease. Bypass in the face of a septic foot or open ulcer will increase the chance of wound and graft infection and the use of heparin or other antithrombotic drugs will increase the chance of wound hemorrhage. Therefore, it is suggested that complications that are specific to the operation or the underlying disease be reported. They fall into local vascular, local nonvascular, and remote and/or systemic categories. In the case of some reported complications, it may be appropriate to separate further those that occur early in the postoperative period from those developing later, with 30 days after operation used as the arbitrary dividing point.

Many complications are difficult to grade in terms of severity other than to identify them as causing death, causing permanent disability, necessitating reoperation, prolonging hospital stay, or as “insignificant.” Table V lists types of complications with a breakdown into subtypes that might be valuable to consider in reports on lower extremity revascularization, as well as suggested grading for outcome or severity.

**Table V.** Types of complications with suggested grading for outcome and severity
Complication (type)	Severity/outcome*
Systemic/remote
Cardiac
Ectopic/arrhythmia	1 = little/no hemodynamic consequence
Congestive failure	2 = symptomatic/required treatment
Myocardial infarction	3 = cardiac arrest/fatal
Stroke/TIA	1 = TIA/temporary deficit
	2 = permanant deficit
	3 = fatal
Deep venous thrombosis	1 = hospitalization not prolonged
Suspected	2 = treatment prolonged hospitalization
Confirmed	3 = required operation
Pulmonary embolism	1 = mild, required antithrombotic drugs
Suspected	2 = serious, required resuscitation
Confirmed	3 = severe, required embolectomy or fatal
Coagulation complications (including drug-induced)
Spontaneous hemorrhage	1 = resolving without treatment
Thrombocytopenia	2 = requiring drug therapy
“White clot syndrome”	3 = requiring operation or fatal
Thrombosis from ATIII, protein C or S deficiency
Renal insufficiency	1 = transient, not requiring dialysis
Contrast media—induced	2 = transient, required dialysis
Thromboembolic	3 = permanent (dialysis, transplant, death)
Ischemic (acute tubular necrosis)
Obstructive
Local/vascular
Graft infection
Early (<30)/late (>30 days)	1 = successful local treatment
Culture positive/negative	2 = required graft removal/bypass
Noninvasive (exposed, contaminated)	3 = loss of limb/life
Invasive, involves graft or anastomoses
Complications of graft/vessel interaction
Intimal hyperplasia (arteriographic, intraoperative, or pathologic diagnosis)	1 = observed, no treatment required
Proximal anastomosis	2 = local treatment sufficed (dilation/revision, local resection)
Distal anastomosis	3 = required “redo” operation
Anastomotic pseudoaneurysm	1 = observed, no treatment required
Mechanical	2 = local treatment sufficed (dilation/revision, local resection)
Infectious	3 = required “redo” operation
Graft complications (exclusive of anastomotic changes)
Dilation/aneurysm	1 = observed, no treatment
Stenosis, focal/diffuse	2 = local treatment sufficed (dilation/revision, local resection)
Elongation/kinking	3 = required “redo” operation
Intrinsic, structural defect†
Arteriosclerotic change†
Technical†
Anastomotic hemorrhage
External bleeding	1 = observed
Internal (hematoma)	2 = required aspiration, drainage
	3 = required anastomotic repair, revision
Graft thrombosis
Early/late	1 = not corrected or corrected with restorative procedure
Cause found	2 = required revision or “redo”
Cause not found	3 = limb loss (unexpected tissue loss)
Unsatisfactory hemodynamic result (despite patency)
Insufficient inflow	1 = > + 1 (but less than expected)‡
Insufficient outflow	2 = + 1‡
“Steal”	3 = < + 1‡
Graft enteric reaction
Anastomotic (fistula) vs. nonanastomotic (erosion)	1 = successfully treated without permanant sequelae
Primary infectious cause vs. no secondary infection	2 = permanant sequelae (e.g., limb loss, -ostomy)
	3 = fatal outcome
Unexpected tissue loss/amputation	1 = minor tissue loss w/o amputation
	2 = minor amputation
	3 = major amputation
Atherothromboembolism	1 = without tissue loss
	2 = with minor tissue loss/amputation
	3 = with major tissue loss/amputation
Colon ischemia	1 = not requiring operation
	2 = colon resection or colostomy
	3 = fatal
Spinal cord ischemia	1 = transient
	2 = minor permanent deficit
	3 = major permanent deficit
Local/nonvascular
Noninfectious wound fluid accumulations
Hematoma	1 = observed, resolved
Seroma	2 = aspirated
Lymphocele	3 = surgical drainage
Wound infections
Superficial	1 = treated with antibiotic
Deep	2 = treated with drainage
Exposed/contaminated graft	3 = required graft removal or bypass
Lymphatic disruption
Lymphedema	1 = no treatment required
Lymphocele	2 = aspiration, drainage
Lymph fistula	3 = exploration with closure of lymphatics
Ureteral injury
Complete obstruction	1 = resolved spontaneously
Partial obstruction	2 = required drainage, diversion
Urinoma (closed leak)	3 = surgical correction or nephrectomy required
Urinary fistula
Sexual dysfunction
Affecting ejaculation (e.g., retrograde)	1 = mild or no effect on sexual activity
Affecting fertility	2 = reduces sexual activity
Affecting erection (potency)	3 = prevents or eliminates sexual activity
Complications of sympathectomy
Disturbance in ejaculation/potency
Neuralgia after sympathectomy
No demonstrable therapeutic benefit
*0 = none; 1 = mild; 2 = moderate; 3 = severe. †These features apply to all subgroups of graft complications exclusive of anastomotic changes. ‡See Criteria for Significant Change in Status, p. 84.

Final comment

TOP

Those who rarely, if at all, publish their results in scientific journals (or even some that do so regularly) may view some of the detailed “demands” encompassed within these recommendations as unnecessarily complicated. Obviously less precision and detail are required in managing data for personal, small group, or society vascular registries. In addition, whereas the broader aspects of this report will have application in these latter situations, the finer details are specifically set forth for those scientific publications that are ultimately intended to influence the practice of vascular surgery. It is hoped that this report will be accepted in that light. We feel confident that most readers will appreciate that precisely defined and uniformly adopted reporting practices will allow us all to better understand the data presented in the articles we read, be able to compare them with reliance, depend more confidently on their conclusions, and base our practices on the more solid foundation of fact and the deeper perspective they will provide.

References

TOP

1. Wibur BG, Olcott C. A comparison of three modes of stress on Doppler ankle pressures. In: Dietrich EB, ed. Noninvasive cardiovascular diagnosis: Current concepts. Baltimore: University Park Press, 1978.

2. Cranley JJ. Ischemic rest pain. Arch Surg 1969;98:187.

MEDLINE

3. Jamieson C. The definition of critical ischaemia of a limb (editorial). Br J Surg 1982;69(Suppl):SI.

4. Yao JST. Hemodynamic studies in peripheral arterial disease. Br J Surg 1970;57:761.

MEDLINE

5. Raines JK, Darling RC, Buth J, Brewster DC, Austen WG. Vascular laboratory criteria for the management of peripheral vascular disease of the lower extremities. Surgery 1976;79:21.

MEDLINE

6.   Siegel ME, Stewart CA. Thallium 201 peripheral perfusion scans: Feasibility of single-dose, single-day, rest and stress study. AJR 1981;136:1179.

7.   DeWeese JA, Reid CG. Autogenous venous grafts five years later. Am Surg 1971;174:346.

8.   Peto R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV, Mantel N, McPherson K, Peto J, Smith PG. Design and analysis of randomized trials required prolonged observations of each patient. I. Introduction and design. Br J Cancer 1976;34:585.

MEDLINE

9. Peto R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV, Mantel N, McPherson K, Peto J, Smith PG. Design and analysis of randomized trials required prolonged observations of each patient. II. Analysis and examples. Br J Cancer 1977;35:1.

MEDLINE

10. Underwood CG, Charlesworth D. Uses and abuses of life table analysis in vascular surgery Br J Surg 1984;71:495.

MEDLINE

11. Colton T. Statistics in medicine, chapter 9. Boston: Little, Brown & Co, 1974.

12. Goldman L, Nussbaum SR, Southwick FS, Krogstad D, Murray B, Burke DS, O'Malley TA, Goroll AH, Caplan CH, Nolan J, Carabello B, Slater EE. Multifactorial index of risk in noncardiac surgical patients. N Engl J Med 1977;297:846.

Publishing and Reprint Information

TOP

Reprint requests: Robert B. Rutherford, M.D., UCHSC, Box 312, 4200 E. Ninth Ave., Denver, CO 80262.
doi:10.1067/mva.1986.avs0040080

Articles with References to this Article

TOP

This article is referenced by these articles:

Basic arterial techniques for peripheral arterial thrombolysis
Techniques in Vascular and Interventional Radiology
June 2001 • Volume 4 • Number 2
Brad M. Hoppenfeld, MD, Jacob Cynamon, MD

ABSTRACT

Use of intravascular ultrasound improves long-term clinical outcome in the endovascular management of atherosclerotic aortoiliac occlusive disease
Journal of Vascular Surgery
April 1998 • Volume 27 • Number 4
Frank Arko, MD, Mark Mettauer, MD, Randy McCollough, MD, Donald Patterson, MD, Larry Manning, MD, Shirley Lee, MSN, RN, Clifford J. Buckley, MD
Temple, Tex.