Download S1936879816304848_mmc1

SUPPLEMENTARY APPENDIX
Supplement to: Prognostic Impact of Chronic Total Occlusions–A report from the Swedish
Coronary Angiography and Angioplasty Registry (SCAAR)
CONTENTS
Participating centers in SCAAR ……………………………………………………… Page 2
Statistics ……………………...……………………………………………………….. Page 3
Missing data in observational studies and rationale for imputation …………….... Page 3
Imputation protocol .……………………………………………………………..... Page 4
Description of the SCAAR/SWEDEHEART registries …………………………….... Page 6
Results ………………………………………………………………………………
Page 11
References …………………………………………….…………………………….... Page 12
Figure legends ...…………………………………….....……………………………... Page 13
Figures ………………………………………………...…………………………….... Page 14
Råmunddal et al: Prognostic impact of CTO /2
PARTICIPATING CENTERS in SCAAR
Borås Hospital, Sweden
Capio, St. Görans Hospital, Sweden
Danderyd University Hospital, Sweden
Eskilstuna Hospital, Sweden
Falun Hospital, Sweden
Gävle Hospital, Sweden
Halmstad Hospital, Sweden
Helsingborg Hospital, Sweden
Jönköping Hospital, Sweden
Kalmar Hospital, Sweden
Karlskrona Hospital, Sweden
Karlstad Hospital, Sweden
Karolinska Solna and Huddinge University Hospitals, Sweden
Kristianstad Hospital, Sweden
Linköping University Hospital, Sweden
Sahlgrenska University Hospital, Sweden
Skövde Hospital, Sweden
Skåne University Hospital, Sweden
Sunderby Hospital, SwedenSundsvall Hospital, Sweden
Södersjukhuset (Stockholm South General Hospital), Sweden
Trollhättan Hospital (NÄL), Sweden
Umeå University Hospital, Sweden
Uppsala University Hospital, Sweden
Västerås Hospital, Sweden
Örebro University Hospital, Sweden
Råmunddal et al: Prognostic impact of CTO /3
METHODS
Statistics
Missing Data in Observational Studies and Rationale for Imputation
Missing data are frequent in observational studies. A common approach in performing
statistical modeling with missing data is to limit analyses to cases with complete data for all
variables in a particular analysis. Such “complete-case” analyses often introduce considerable
bias and are always inefficient. Bias arises if individuals with missing data are not
representative of the population of interest. Inefficiency is caused by reduced sample size,
which decreases statistical power (e.g., by decreasing the number of events).
The literature about missing data in statistical modeling is extensive.1, 2 Ad hoc imputation
methods such as “the last observation carried forward”, insertion of a “missing category
indicator”, and imputation protocols in which each missing value is replaced with “an
assumed or estimated value” are not recommended, as they often lead to reduction or
exaggeration of the association of interest.3 Instead, more accurate imputation methods have
been developed based on classification of missing data mechanisms and on probability
models4-6.
Data are missing completely at random if the probability that a particular observation is
missing does not depend on observable variables. Data are missing at random if the
probability that observations are missing is independent of the missing data. Data are missing
not at random if the probability of missing still depends on the missing value even after the
available data are taken into account. When data are missing not at random, valid inferences
require explicit assumptions about the mechanisms that led to missing data. Missing data
gathered in health care registries are often missing at random. Methods to deal with data
missing at random fall into three principal categories: likelihood-based approaches, weighted
estimation, and multiple imputation. Multiple imputation is the most commonly used because
of flexibility particularly when multiple variables have missing values. The general
recommendations about how to report and handle missing data in observational studies is
Råmunddal et al: Prognostic impact of CTO /4
provided in the consensus document Strengthening the Reporting of Observational Studies in
Epidemiology (STROBE).
Imputation protocol
We examined our database for missing data. Logistic regression showed that a number of
variables were associated (P<0.05) with missing data, including diabetes, previous myocardial
infarction (MI), hyperlipidemia, hypertension, smoking habits, and other. This relationship
indicates that the presence of missing data was not completely random. Thus, in addition to
the complete case analysis, we applied multiple imputation method to estimate the missing
data.5, 6
For our primary analysis, we used Cox proportional-hazards regression based on the
imputed data set under the assumption that missing data are missing at random. The
imputation protocol consisted of the chain-equation method7 and a predictive-mean matching
algorithm. We used the same covariates in the imputation protocol as in the main analysis
with addition of Nelson Aalen cumulative hazard and event indicator. We imputed 20 data
sets with 100 cycles between the outputs. The adequacy of the iterations (convergence) was
verified by visual inspection of trace plots for different chains, which showed no apparent
trends for the imputed values between the imputation models. Continuous variables were
imputed by ordinary least-squares regression, whereas binary variables were imputed by
logistic regression and categorical variables by multinomial logistic regression. The following
variables had missing data and were imputed: age, diabetes, hypertension, hyperlipidemia,
smoking status, previous infarction, previous PCI, complication, hospital category, primary
decision. The following variables were included in the imputation model as regular variables:
severity of coronary artery disease, gender, indication for intervention, year of intervention,
treated vessel, arterial access site indicator of missing data and vital status. We included
Nelson-Aalen cumulative hazards indicator in addition to above-mentioned variables for
imputation algorithm because its inclusion decreases the bias in survival data8. The imputation
Råmunddal et al: Prognostic impact of CTO /5
procedure and subsequent analyses were performed according to the Rubin’s protocol5 using
Stata software9 (version 13.1, StataCorp, College Station, TX). The following imputation
models were used for each variable with missing data:
Age (linear regression): hypertension, hyperlipidemia, diabetes, previous infarction, previous
PCI, complication, hospital category, smoking status, primary decision, severity of coronary
artery disease, gender, indication for intervention, year of intervention, treated vessel, NelsonAalen cumulative hazard, indicator of missing data, vital status.
Hypertension (logistic regression): age, hyperlipidemia, diabetes, previous infarction,
previous PCI, complication, hospital category, smoking status, primary decision, severity of
coronary artery disease, gender, indication for intervention, year of intervention, treated
vessel, Nelson -Aalen cumulative hazard, indicator of missing data, vital status.
Hyperlipidemia (logistic regression): age, hypertension, diabetes, previous infarction,
previous PCI, complication, hospital category, smoking status, primary decision, severity of
coronary artery disease, gender, indication for intervention, year of intervention, treated
vessel, Nelson-Aalen cumulative hazard, indicator of missing data, vital status.
Diabetes (logistic regression): age, hypertension, hyperlipidemia, previous infarction,
previous PCI, complication, hospital category, smoking status, primary decision, severity of
coronary artery disease, gender, indication for intervention, year of intervention, treated
vessel, Nelson-Aalen cumulative hazard, indicator of missing data, vital status.
Previous infarction (logistic regression): age, hypertension, hyperlipidemia, diabetes,
previous PCI, complication, hospital category, smoking status, primary decision, severity of
coronary artery disease, gender, indication for intervention, year of intervention, treated
vessel, Nelson-Aalen cumulative hazard, indicator of missing data, vital status.
Previous PCI (logistic regression): age, hypertension, hyperlipidemia, diabetes, previous
infarction, complication, hospital category, smoking status, primary decision, severity of
coronary artery disease, gender, indication for intervention, year of intervention, treated
vessel, Nelson-Aalen cumulative hazard, indicator of missing data, vital status.
Råmunddal et al: Prognostic impact of CTO /6
Complication (logistic regression): age, hypertension, hyperlipidemia, diabetes, previous
infarction, previous PCI, hospital category, smoking status, primary decision, severity of
coronary artery disease, gender, indication for intervention, year of intervention, treated
vessel, Nelson-Aalen cumulative hazard, indicator of missing data, vital status.
Hospital category (multinomial logistic regression): age, hypertension, hyperlipidemia,
diabetes, previous infarction, previous PCI, hospital category, smoking status, primary
decision, severity of coronary artery disease, gender, indication for intervention, year of
intervention, treated vessel, Nelson-Aalen cumulative hazard, indicator of missing data, vital
status.
Smoking status (multinomial logistic regression): age, hypertension, hyperlipidemia, diabetes,
previous infarction, previous PCI, hospital category, primary decision, severity of coronary
artery disease, gender, indication for intervention, year of intervention, treated vessel, NelsonAalen cumulative hazard, indicator of missing data, vital status.
Description of the SCAAR/SWEDEHEART Registries
The SWEDEHEART Registry
The information about The Swedish Web-system for Enhancement and Development of
Evidence-based care in Heart disease Evaluated According to Recommended Therapies
(SWEDEHEART) was published previously10. SWEDEHEART was established in December
2009 after merging of the National Registry of Acute Cardiac Care (RIKS-HIA), the Swedish
coronary angiography and angioplasty registry (SCAAR), the Swedish Heart Surgery
Registry, and the National Registry of Secondary Prevention (SEPHIA). RIKS-HIA was
developed in 1990 and was established as a national quality registry in 1995. SEPHIA was
added to RIKS-HIA in 2005 to register effects of secondary prevention efforts in patients with
acute MI. SCAAR was established in 1998 after a merge of the Swedish National Angioplasty
Registry and the Swedish National Coronary Angiography Registry, both initiated in the early
Råmunddal et al: Prognostic impact of CTO /7
1990s by hospitals which at that time performed coronary angiographies and PCIs. The
Swedish Heart Surgery Registry was formed in 1992.
Organization and Funding
SWEDEHEART is managed by a steering group, consisting of the chairmen of the working
groups of the registries and representatives from the Swedish Heart Association and the
Swedish Society of Cardiac Nursing. Uppsala Clinical Research Center (UCR) has developed
the web-based version of the registry and is responsible for project management,
administration, monitoring, quality controls, and statistical reports. The registry is financed by
the Swedish Association of Local Authorities and Regions (the public health care provider),
and is supported by the Swedish Heart Association, the National Board of Health and
Welfare, and the Swedish Heart and Lung Foundation. Participating hospitals are not
reimbursed by SWEDEHEART, and costs of local data entry are covered by internal budgets.
Data
SWEDEHEART includes patients admitted to hospital because of symptoms of an acute
coronary syndrome (ACS) and patients who undergo coronary angiography/angioplasty or
heart surgery. The registry enrolls approximately 80,000 cases each year: 30,000 with ACS,
40,000 undergoing coronary angiography or angioplasty, 7,000 undergoing heart surgery, and
6,000 who are followed for 12–14 months for secondary prevention after an ACS. The
registry is web-based. All data are registered on-line directly by the caregiver and transferred
in encrypted form to a central server. During registration, the whole process of care is kept
together in one record even if the patient is transferred between different units and hospitals.
The technical platform, OpenQreg, is published as open-source software that can receive data
via the Internet or from other databases and electronic patient journals. The platform is in
direct contact with the Swedish National Population Registry for immediate access to personal
data and deaths.
Råmunddal et al: Prognostic impact of CTO /8
For patients admitted to hospital because of symptoms of an ACS, information on 106
variables is collected prospectively, including patient demographics, admission logistics, risk
factors, medical history, medical treatment before admission, electrocardiographic changes,
biochemical markers, other clinical features and investigations, medical treatment in hospital,
interventions, hospital outcome, and diagnoses and medications at discharge. Patients younger
than 75 years of age who have been hospitalized for ACS return for follow-up visits 6–10
weeks and 12–14 months after discharge. From these visits, approximately 75 new variables
are added.
For patients undergoing coronary angiography/angioplasty for any clinical indication,
approximately 150 variables are registered. Besides baseline characteristics, the registry
includes a detailed description of angiographic findings, procedures, type of stenosis, type of
stent, antithrombotic treatment, and complications. The system has an interactive method for
registration of restenosis and stent thrombosis. Detailed information about every previously
implanted stent anywhere in the country is presented and a mandatory question about
existence of any form of restenosis or stent thrombosis has to be answered. Every hospital in
Sweden providing the relevant services participates in the SWEDEHEART registry.
Patient Identification
Every Swedish citizen has a unique personal identification number which together with name,
address, and hospital identity is included in the registry. The use of personal identification
numbers enables merging of the SWEDEHEART database with the National Cause of Death
Register, which includes information about the vital status of all Swedish citizens, and the
National Patient Registry, which includes diagnoses at discharge for all hospital stays in
Sweden. All patients are informed about their participation in the registry and the follow up,
and have the right to decline participation. Every merge of registries is approved by the
National Board of Health and Welfare, the Swedish Data Inspection Board, and the ethical
Råmunddal et al: Prognostic impact of CTO /9
committee at Uppsala University. After merging of the registries, researchers have access to
hospital identity but not to patient identity.
Data Quality
Uppsala Clinical Research Center provides manuals, education, and technical advice,
including a telephone help desk for all users of the registry. The system has error-checking
routines for range and consistency. Definitions are easily available when data are entered. To
ensure the correctness of the data entered, a monitor visits about 20 hospitals and compares
data entered into the SWEDEHEART with the information in the records of 30–40 randomly
chosen patients in each hospital. When 637 randomly chosen computer forms from 21
hospitals containing 38,121 variables were reviewed in 2007, there was 96.1% agreement
(range: 92.6–97.4%). To reach a high degree of completeness, a majority of variables are
mandatory and each hospital can monitor data completeness. The system provides all users
with an array of on-line interactive reports regarding changes in processes of care and
outcome in direct comparison with other hospitals. The SCAAR registry also works as a
clinical tool, as it displays detailed information about any previously performed intervention.
After a coronary procedure, two reports summarizing the findings and intervention performed
are printed: one for the patient and one for the patient’s clinical files. The second part includes
a request to the ward to report potential complications after PCI. The registry captures 100%
of patients undergoing angiography, angioplasty, or heart surgery. Regarding patients with
ACS, 60% are captured by the registry; the percentage varies considerably between hospitals.
The main reason for this is that some ACS patients are admitted to other units than coronary
care units. The degree of patient capture is higher in younger patients and in those with STelevation MI (STEMI).
Use of SWEDEHEART Data
The main purpose of the registry is to support the improvement of care and the development
of evidence-based therapy for coronary artery disease by providing continuous information on
Råmunddal et al: Prognostic impact of CTO /10
care needs, therapy, results of therapy, and changes within a hospital as well as in comparison
to other hospitals. The long-term goals are to contribute to decreased mortality and morbidity
among patients and to increase the cost effectiveness of coronary care. A national, regional,
and county-based report is presented on a yearly basis concerning a large number of variables
that is open for public. The registry compares performance of participating hospitals and
different treatment modalities and medical devices. The results, especially regarding
differences between hospitals and adherence to national guidelines, have been discussed in
different media and by authorities—further contributing to improvements in care. In addition,
many hospitals are engaged in collaborations on quality-development projects, which are
supported by the on-line interactive reporting system as a continuous quality-control
instrument. By giving each hospital an opportunity to compare its treatments and results over
time and with other hospitals, the registry has been a powerful tool for improvements both
locally and nationally. SWEDEHEART and its original four registries have, so far, been the
source of more than 100 original scientific papers, several of which were published in highranking journals.
Randomized Registry Clinical Trials (RRCT) in SWEDEHEART
SWEDEHEART recently introduced the concept of a randomized registry clinical trials
(RRCT). The rationale for the RRCT concept is that both randomization (to a specific
treatment/intervention) and follow-up are conducted within the registry. In the TASTE trial10
(Thrombus Aspiration in ST-elevation in Scandinavia), 7200 patients were included to test the
effect of thrombus aspiration on mortality in STEMI. The most important advantage of the
RRCT (particularly with a large number of participating hospitals) over traditional
randomized clinical trials is the capacity to include numerous patients over a short time in a
cost-efficient way. Other RRCTs are currently ongoing within the SWEDHEART, such as
DETOX (oxygen vs. air in STEMI) and VALIDATE (bivalirudin vs. heparin in NSTEMI and
STEMI).
Råmunddal et al: Prognostic impact of CTO /11
Reporting of CTOs in SCAAR
There are two methods for on-line reporting CTO patients in SCAAR. The first method,
introduced in 2005, is based on the information about the percentage of luminal stenosis at the
level of coronary segments (Figure 1A). From 2005 onward, the information derived from a
diagnostic coronary angiogram can also be used to determine whether a coronary segment was
totally occluded. The operator reports the grade of lumen narrowing (0–100%), whether
occlusions is older than 3 months, and whether vessel diameter >2 mm. The second method is
based on the separate variable by which PCI operators classify a treated occlusion either as a
chronic occlusion > 3 months’ duration or as an acute/subacute occlusion < 3 months’
duration (Figure 1B).
Validation
The definition of CTO was validated in a subgroup of 955 patients from one university
hospital (Sahlgrenska University Hospital) and three county hospitals (Norra Älvsborgs
Hospital, Borås Hospital, Skövde Hospital). This subgroup included 5.7% of all identified
CTO patients in SCAAR during the study period. The patients were randomly selected with a
random number generator. The validation was done by a panel of five experienced
interventional cardiologists, who examined individual coronary angiograms according to a
predefined monitoring plan. Each angiogram was evaluated to determine whether the patient
had a previous CABG, whether the treated occlusion was ≥3 months old, and whether 100%
segmental stenosis on the angiogram was caused by an occlusion ≥3 months old. The results
of the validation were then compared to the data entered in SCAAR. SCAAR data were not
validated or adjudicated by an independent core lab.
RESULTS
Patients and Procedures
During the study period, 276,931 coronary angiographies and/or PCI procedures in 215,836
patients were performed in Sweden. Complete information about percent stenosis in coronary
Råmunddal et al: Prognostic impact of CTO /12
segments was available in 160,159 (57.8%) angiographies, of which 144,744 were in patients
without previous CABG. In total, 89,872 patients without previous CABG had >50% luminal
stenosis on angiography, and 14,411 had a CTO. There were 75,431 non-CTO patients. CTO
patients were more often male and were more likely to have diabetes, hypertension,
hyperlipidemia, previous MI, and previous PCI (Table 1). The extent of CAD was also more
severe in CTO patients. CTO was diagnosed in the majority of cases during coronary
angiography for stable CAD; however, 41.5% of CTOs were in patients with ACS. The CTO
was located most often in the RCA (42%), followed by the LAD (23%) and the LCx (21%).
In approximately 14% of patients, CTOs were present in more than one vessel. CTOs were
more frequent in proximal segments (64.1%) than in distal segments (35.9%). Complete Ad
hoc PCI was performed in
Validation
The validation procedure disclosed 36 (3.8%) erroneously classified patients. Of these, 18
patients did not have an occluded coronary artery on the coronary angiogram, 5 had a prior
CABG, and 13 had acute coronary occlusions. All acute occlusions were classified correctly.
CTO and Long-Term Mortality
For sensitivity analysis of our primary model we used propensity score (PS) adjustment. We
calculated generalized PS for each patient using logistic regression according to Hirano et al.11
using dedicated Stata package12. For the PS stratification analyses, strata were created based
on PS quintiles. Formation of five strata based on the PS values removes >90 % of the bias in
each of the included covariates in the propensity model13. The following covariates were
entered into the multivariable logistic regression regression: age, gender, body mass index,
hypertension, diabetes mellitus, tobacco use, hyperlipidemia, previous myocardial infarction
(MI), previous PCI, previous CABG, indication for coronary angiography, severity of
coronary artery disease, hospital volume of coronary angiography and PCI, year of procedure,
puncture site, complications and primary decision after diagnostic angiography (medical
Råmunddal et al: Prognostic impact of CTO /13
treatment, PCI, CABG). The calculated PS was entered into the Cox proportional-hazards
regression as quintiles of PS. We assessed goodness-of-fit (calibration) for the models was
using the Hosmer-Lemeshow test. The discriminative ability of the model was assessed using
the c-statistic. Estimated mortality risk associated with the presence of CTO based on
adjustment with PS (HR 1.34; 95% CI 1.28 - 1.41) was similar to the risk estimation based on
the main model (HR 1.29; 95% CI 1.22 - 1.37).
CTO in Patient Subgroups
Severity of CAD
Complete case analysis showed similar results as the primary model. Multivessel disease was
associated with increased mortality compared to one vessel disease (two-vessel: HR 1.14,
95% CI 1.08–1.20, P<0.001; three-vessel: HR 1.34, 95% CI 1.26–1.41, P<0.001; left main:
HR 1.57, 95% CI 1.47–1.69, <0.001). There was no significant interaction between CTO and
severity of CAD in complete case analysis.
Age
Complete case analysis showed similar results as the main model (HR 0.98, 95% CI 0.98–
0.99, P<0.001) for every year of increased age in the presence of a CTO. Complete case
analysis with age categorized into four groups showed similar results as the main model (age
<59: HR 1.64, 95% CI 1.38–1.94, P<0.001; age 60–69: HR 1.48, 95% CI 1.32–1.65,
P<0.001; age 70–79: HR 1.23, 95% CI 1.13–1.33, P<0.001; age >80: HR 1.13, 95% CI 1.03–
1.25, P<0.001).
Gender
Complete case analysis showed similar results as the main model (men: HR 1.25, 95% CI
1.18–1.34, P<0.001; women: HR 1.21, 95% CI 1.09–1.34, P<0.001).
Diabetes
Råmunddal et al: Prognostic impact of CTO /14
Complete case analysis showed similar results as the main model (diabetes: HR 1.26, 95% CI
1.16–1.38, no diabetes HR 1.30, 95% CI 1.22–1.38, P<0.001; P<0.001).
Råmunddal et al: Prognostic impact of CTO /15
1.
2007.
References
McKnight PE. Missing data : a gentle introduction. New York ; London: Guilford;
2.
Molenberghs G, Kenward MG. Missing data in clinical studies. Chichester: Wiley;
2007.
3.
Horton NJ, Kleinman KP. Much ado about nothing: A comparison of missing data
methods and software to fit incomplete data regression models. Am Stat 2007;61(1):79-90.
4.
23.
Little RJ. RD. Statistical Analysis with Missing Data. In. New York: Wiley; 2002, 19-
5.
Rubin DB. Inference and Missing Data. Biometrika 1976;63(3):581-590.
6.
Rubin DB. Multiple imputation for nonresponse in surveys. Hoboken, N.J.: WileyInterscience; 2004.
7.
van Buuren S. Multiple imputation of discrete and continuous data by fully conditional
specification. Statistical methods in medical research 2007;16(3):219-42.
8.
White IR, Royston P. Imputing missing covariate values for the Cox model. Statistics in
medicine 2009;28(15):1982-98.
9.
White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues
and guidance for practice. Statistics in medicine 2011;30(4):377-99.
10. Frobert O, Lagerqvist B, Olivecrona GK, Omerovic E, Gudnason T, Maeng M, Aasa M,
Angeras O, Calais F, Danielewicz M, Erlinge D, Hellsten L, Jensen U, Johansson AC,
Karegren A, Nilsson J, Robertson L, Sandhall L, Sjogren I, Ostlund O, Harnek J, James SK,
Trial T. Thrombus aspiration during ST-segment elevation myocardial infarction. The New
England journal of medicine 2013;369(17):1587-97.
11. Hirano KI, G. W. The propensity score with continuous treatments. In Applied Bayesian
Modeling and Causal Inference from Incomplete-Data Perspectives 2004:73–84.
12. Bia M, Mattei A. A Stata package for the estimation of the dose-response function
through adjustment for the generalized propensity score. Stata J 2008;8(3):354-373.
13. Shah BR, Laupacis A, Hux JE, Austin PC. Propensity score methods gave similar
results to traditional regression modeling in observational studies: a systematic review. J Clin
Epidemiol 2005;58(6):550-559.
Råmunddal et al: Prognostic impact of CTO /16
Figure Legend
The figure represents the screenshot from the SCAAR web interface in Swedish and shows
how data about CTOs are reported in the registry. After diagnostic angiography, the operator
reports the percentage of luminal stenosis for each coronary segment (Figure S1-A). The
percentage of luminal stenosis is divided into six categories; 0–29%, 30–49%, 50–69%, 70–
89%, 90–99%, and 100% (red arrow). When 100% luminal stenosis is entered, the operator is
asked to answer whether the lesion is a CTO (defined as occlusion >3 months old). The
operator is also asked whether the occluded segment is >2 mm or <2 mm in diameter (red
rectangle).
After every PCI, the operator report data about the procedural details at the segmental level
(Figure S1-B). Information is collected about intervention-related details and lesion
characteristics. This includes data about lesion type (de novo, in-stent restenosis, other),
native vessel or graft, stent name, stent length, stent diameter, post-dilatation of stent, balloon
diameter, balloon pressure, local success, the use of diagnostics including FFR (fractional
flow reserve), IFR (instantaneous wave-free ratio), IVUS (intravascular ultrasound), OCT
(optical coherence tomography). The results of the diagnostic procedures are reported to a
separate module. If the treated segment was an occlusion, the operator documents whether the
occlusion was > 3 or < 3 months old (red rectangle).
Råmunddal et al: Prognostic impact of CTO /17
Figure S1-A
Agiografiskt fynd=angiographic findings, Vengraft=vein graft, Artärgraft=arterial graft,
Normalt fynd=normal finding; Beräkna fynd=calculate findings, Fynd=fidings; 3-kärl ej HS=
3-vessel without left main, ja=yes; Segment större än 2mm=segment larger than 2 mm.
Råmunddal et al: Prognostic impact of CTO /18
Figure S1-B
Segmentnummer=segment number, Graft=graft, Nummer på stenos i samma segment=
number of stenosis in same segment (Första=first); Ocklusion=occlusion, Stenostyp=type of
stenosis (de novo, in-stent or other), Stenosklass= classification of the stenosis (type A, type
B1, type B2, type ), Procedurtyp=type of procedure (direct stenting, balloon dilatation,
balloon dilation + stent, drug-eluting balloon, diagnostic), stent=name of stent and name of
stent manufacturer, Maxtryck ballong i stent= maximal pressure in stent balloon, Stentlängd
(mm)=length of stent in millimeters, Stentslut=end of stent (which segment), Eftedilatation av
stent=after-dilatation of stent, Lokal framgång=local success
Råmunddal et al: Prognostic impact of CTO /19
Figure S2