S. L u g a u e r , A . R e g e n f u s , M . B 6 s w a l d , E M a r t u s , C. G e i s , T. B e c h e r t , 3. G r e i l , J.-P. G u g g e n b i c h l e r
A New Scoring System for the Clinical Diagnosis of Catheter-Related Infections Summary: It is difficult to make the clinical diagnosis of catheter-related infections using the available and established definitions of the HICPAC (Hospital Infection Control Practices Advisory Committee) of the CDC (Centers for Disease Control, definitions of nosocomial infections). The scoring system shown here is a modification of these definitions and has enabled the causal relationship between the catheter and clinical episodes of systemic infections to be quantitatively graded. The scoring system included the following criteria: height and rate of rise of body temperature, attendant shivering, identification of pathogens in blood and/or catheter tip cultures, improvement in the clinical course alter catheter removal, signs of catheter exit site inflammation and results of diagnostic tests for other possible sources of infection. These criteria were graded using points and weighted according to their specificity. The comparative evaluation of 65 episodes of systemic infections using the scoring system and the diagnostic criteria of HICPAC showed agreement in 85%. No case was graded "false-negative." In nine of ten false-positive cases additional findings supported the presence of a catheter-associated infection. This scoring system appears, therefore, to be more sensitive than existing diagnostic criteria, without loss of specificity.
Introduction Two definitions are currently used for the diagnosis of a central venous catheter (CVC)-related infection. 1. The criteria used in the current guidelines of the H 1CPAC (Hospital Infection Control Practices Advisory Committee) of the C D C (Centers for Disease Control) of 1996 (Table 1 [1 ]).These criteria are mainly microbiological and arc based on the isolation of the same microbe from microbiological cultures of the surface of the catheter and of the blood. Because of their relatively high specificity, these criteria are suitable for studies in which the causal relationship between the catheter and clinical signs is important. 2. The definition of (catheter-related) primary scpsis, as applied by the national reference center for hospital infection control at the Robert Koch Institute in accordance with the C D C criteria for nosocomial infections (Table 2 [2]). These definitions focus on the epidemiological evaluation of the incidence and prevalence of CVCrelated infections. The association between the C V C and primary sepsis is given by the concurrence of CVC use and the absence of any other apparent source of infection, without diagnostic confirmation of this causality. On the basis of these definitions, the clinical diagnosis of CVC-related infections in patients with multiple pathology is difficult in the case of 9 other loci of infection (e. g. pneumonia) causing similar clinical signs, 9 other invasive biomaterials (ventilation tubes, Sheldon catheter, drainage etc.), as possible additional sources of infection, 9 false-negative results of blood cultures, possibly as a result of concomitant antimicrobial therapy,
9 false-negative culture results on the catheter surface possibly caused by antibiotics given about the time of the blood culture or the removal of CVC, 9 clinical i m p r o v e m e n t after initiation or change of antibiotic therapy suggesting a positive time correlation with catheter removal. A further disadvantage of definitions based mainly on the results of microbial cultures is the low sensitivity and predictive value of blood cultures, if only one culture is investigated, or - as is often the case - no microbiological tests are routinely done.
Materials and Methods On the basis of the above definitions we established a scoring system (Table 3) which included the following criteria: height and rate of rise of body temperature, attendant shivering, identification of pathogens in blood and/or catheter tip cultures, improvement of the clinical course after catheter removal, signs of catheter exit site inflammation and results of diagnostic tests for other possible sources of infection. These criteria were graded using points on the basis of defined conditions (e. g. concomitant therapy with antibiotics) in welldefined time relations. For this, data on patients and catheters, the time course of medication, blood cultures, clinical signs and other diagnostic tests were evaluated separately using the indiDr. recd. X Lugauer. I)r. mcd. Anja Regenfi~s, l)r. mcd. M. BOswald, l)r. rer. nat. Z Bechert, Dr. reed. Z Greil. Prof. Dr. reed. J.-P Guggenbichler, Klinik mit Poliklinik ftir Kinder und Jugcndliche der FricdrichAlexander-Univcrsit~it Erlangen-NiJrnbcrg, Loschgestr. 15, 1)-91054 Erlangen; Dr. rcr. nat./-~ Martu.s; Dipl.-Biol. Christine Geis, Institut for Medizinischc Statistik und I)okumcntation, Fricdrich-AlcxanderUniversitat Erlangcn-Ntirnberg, Waldstr. 6, I)-91054 Erlangen, Germany.
Infection 27 (1999) Suppl. 1 9 MMV Medien & Medizin VcrlagsGmbH Mtinchcn 1999
S 49
S, Lugauer et al.: Catheter-Associated Infection Score System
Table 1: Definitions of catheter-related infection as recommended by CDC/HICPAC (Centers for Disease Control/Hos)ital Infection Control Practices Advisory Committee) 1996. Catheter colonization
Growth of > = 15 colony-forming units (semiquantitative culture) 3 or > 10 (quantitatwe culture) from a proximal or distal catheter segment in the absence of accompanying clinical symptoms
Exit-site infection
Erythema, tenderness, induration, or purulence within 2 cm of the skin at the exit site of the catheter
Pocket infection
Erythema and necrosis of the skin over the reservoir of a totally implantable device, or purulent exudate in the subcutaneous pocket containing the reservoir
Tunnel infection
Erythema, tenderness and induration in the tissues overlying the catheter arid > 2 cm from the exit site
Catheter-related bloodstream infection (CR-BSI)
Isolation of the same microorganism (i. e. identical species, antibiogram) from a scmiquantitative or quantitative culture of a catheter segment and from the blood (preferably drawn from a peripheral vein) of a patient with accompanying clinical symptoms of BSI and no other apparent source of infection
In the absence of Defervescence after removal of an laboratory confirmation implicated catheter from a patient with BSI may be considered indirect evidence of CR-BSI Infusion-related bloodstream infection
Isolation of the same microorganism from infusion fluid and from separate percutaneous blood cultures, with no other identifiable source of infection
vidual patient's records. Data were connected using a relational database as shown above [3]. The definitions used were as follows: Fever - m a x i m u m : Highest temperature noted within evaluated infectious episode. Fever - spiking: Rate of rise of body temperature > 1 ~ h. Fever - shivering." Any documented chills. Clinical improvement after catheter removal: A positive time correlation was defined as defervescence or return of symptoms to normal the same day or within the 2 days following CVC removal. Any new antibiotic or change of antibiotics within 2 days before convalescence were taken into consideration. B l o o d cuhure: All results of blood cultures taken (peripheral of central) during CVC use were accepted, if the same microorganism was identified in the blood and on the catheter surface, or Staphylococcus epidermidis or Staphylococcus aureus, regardless of CVC culture, were idcntified. Any antibiotic therapy on the day of blood culture was taken into consideration.
S 50
Table 2: Definition of primary sepsis according to the CDC criteria for nosocomial infections. Laboratory confirmed I. Isolation of microorganism in blood culture that are not associated with another localized infection or 2. One of the following: Fever > 38~ shivering or hypotension and one of the following: 9 common skin-associated pathogenic organism isolated in two blood cultures drawn at different times and not associated with another localized infection, 9 common skin-associated pathogenic organism isolated in blood culture of a patient with intravascular device and physician initiates antibiotic therapy, 9 positive antigen test and pathogen is not associated with infection at another locus. Clinical sepsis Fever (>38~ or hypotension (systolic BP < 90 mmHg) or oliguria (< 20 ml/h) without other known cause and all of the following: 1. no blood culture done, no microorganism isolated or no positive antigen test, 2. no apparent localizcd infection elsewhere, 3. physician initiated antibiotic therapy because of clinical suspicion of sepsis. Catheter insertion site: Any documented sign of redness, swelling, warmth, secretion at the exit site the same day or within 5 days prior to CVC removal. Catheter cultures: Bacterial growth after enrichment in roll-out, flushing (both quantitative) or broth (qualitative) culture of the CVC tip. Any antibiotic therapy on the day of catheter removal was considered. Findings in other foci: Any positive in X-ray examinations, urine cultures, CSF, BAL, signs of wound infection, laboratory findings, cultures of other invasive devices within a period of 7 days before the onset of symptoms and 1 day after resolution of symptoms. Additionally, a statement was requested from the primary physician. The absence of positive results in other foci despite full investigations was defined as the performance relevant investigations, the results of which were all negative. Cases with no documented investigations were considered as "not evaluable," even if the primary physician found no other focus of infection.
am 9
~
9
i
0
0~
000~
~ g ~@~ @
Diagnosis | discrepant
0
7
16 sum of score points
Figure 1: Association of score sum using the scoring system and the results using the HICPAC criteria (see Table 4).
Infection 27 (1999) Suppl. 1 9 MMV Medien & Mcdizin VerlagsGmbH Mtinchen 1999
S. Lugauer et al.: Catheter-Associated Infection Score System
Table 3: SCoring System for the quantitative evaluation of suspected clinical catheter-related infections (CRI).
No 0 < 38.5 ~ 1 38.5 < 39.5 ~ 2 >= 39.5 ~ 3 Fever - spiking No 0 Not evaluable 1 Yes 2 Fever - shivering No 0 Not evaluable 1 Yes 2 Improvement after No/no time 0 catheter removal correlation Time correlation + new or change of antibiotics 1 Time correlation/no antibiotics 2 Blood culture Sterile 0 Not done 1 Sterile during antibiotic therapy 2 Positive 3 Catheter insertion No inflammation 0 site within 5 d before Not evaluable 1 removal Inflammation 2 Catheter cultures Sterile, no antibiotic therapy 0 Not done 1 Sterile during antibiotic therapy 2 Positive in broth enrichment culture only 2 Scanty (< 15 CFU) 3 Moderate numbers (15 - 100 CFU) 4 Numerous (> 100 CFU) 5 Findings in other foci Yes 0 Not evaluable 1 No, despite relevant investigations 2 Definitely related 24 > < 33 Catheter related Probably related 16 > < 24 Possibly related 8 > < 16
tient's case history were evaluated and documented by three trained physicians (AR, BS, AB) in three different hospitals.The HICPAC criteria were used as the gold standard. Discrepant results by the two systems were given a further case review.
Fever - maximum
Unlikely related
0>< 8
x 0.5
x 0.25
x 0.25
x3
x2
xl
x2
x3
Not catheter related
Following this grading, all criteria were weighted using a multiplication factor. Using a cut-off value of 16.0 points, infectious episodes were classified as catheter related or as infections with other causes. Based on the data of 65 patients with clinical signs of systemic infection, we compared the results of this scoring system with those using the HICPAC criteria. These data based on the pa-
Results Accoring to H I C P A C criteria, 20 of the 65 episodes of systemic infections were defined as catheter related, compared to 30 using the scoring system. In 55 cases (85%) the results were in agreement (Table 4). There were no false-negative cases, while ten episodes were assessed as false-positive in contradiction to the H I C P A C criteria. These cases are shown in detail in Table 5. Of seven cases with nonsignificant catheter colonization, the number of C F U was low in three cases and in an additional three cases patients had been given systemic antibiotic therapy at the time of catheter removal. In six cases, two additional criteria (positive blood culture, improvement after catheter removal, signs of exit site inflammation or no other apparent infectious source) supported the causal relationship with the catheter. Three false-positive cases were defined as not related according to the H I C P A C criteria because of another apparent site of infection. In one case blood cultures were negative, in another they were not done and in the third they were sterile during antibiotic therapy. The clinical course in all three cases improved after catheter removal. In one case there was additional evidence of exit-site inflammation. In conclusion, the results in nine of the ten discrepant cases supported a causal connection with the catheter. The hypothesis that these cases were defined as false-negative according to the H I C P A C criteria should be discussed. Only one case lacked any additional evidence of a positive CVC relationship in the absence of another infectious focus. This case scored 16.0 points, which was the precise cut-off. The graphical comparison of the 'individual score sums and the results according the H I C P A C criteria showed a positive quantitative association (Figure ] ). All discrepant cases were located in the upper range of the positive side.
Discussion This scoring system does not contradict the existing definitions of the H I C P A C and C D C but modifies and enhances them. The data show this scoring system to have improved sensitivity without loss of specificity. This score provides a quantitatively graded evaluation. Well-defined criteria made a centralized evaluation of multicentric clinical data possible. A n effective and reliable result was achieved even if some data were not available or had not been documented. The evaluation of the microbial culture, taking antibiotic therapy into account, may reduce the risk of a wrong diagnosis because of false-negative cultures. In addition
Infection 27 (1999) Suppl. 1 9 MMV Medien & Medizin VerlagsGmbH Manchen 1999
S 51
S. Lugauer et al.: Catheter-Associated Infection Score System
Table 4: Comparison of the results of scoring and from application of the HICPAC criteria (1996) in 65 systemic infectious episodes.
I
Not catheter related
7
19
7
-
33
Local exit site infection
1
6
-
-
7
III
Colonized catheter and other localized infection
-
2
3
-
5
IV
Colonized catheter without other localized infection
-
-
4
4
8
V
Catheter-related sepsis (clinical diagnosis)
-
-
1
-
1
VI
Catheter-related sepsis (laboratory confirmed)
II
Z
-
-
3
8
11
8
27
18
12
65
Clinicalsignsof systemicinfectionplus; I sterileor missingcatheter tip culture; II I + localexit-siteinflammation; III significantlycolonizedcatheter tip and other localizedinfection; IV significantlycolonizedcatheter tip withoutother localizedinfection; V positivebloodcultureand defervescenceafter removalof catheter; VI isolationof the identicalmicroorganismat catheter tip and in bloodculture.
Table 5: Case review of false-positive cases (positively catheter related according to the scoring system, but negative according to the HICPAC criteria).
HICPAC: colonized catheter + sign of systemic infection with other localized infection 1
> 15 CFU
n.d.
Yes, ab (+)
Inflammation
Yes
18.0
2
> 15 CFU
Sterile, ab (-)
Yes, ab (+)
Not inflamed
Yes
18.5
3
> 15 CFU
Sterile, ab (+)
Yes, ab (+)
Unknown
Yes
18.5 21,0
HICPAC: infection not catheter related
*
4
Broth enrichment only
Positive
Yes, ab (-)
Unknown
None
5
< 15 CFU
n.d.
Yes, ab (-)
Not inflamed
Unknown
19.5
6
Sterile during ab
Positive
No
Unknown
None
19.0
7
Sterile without ab
Positive
Yes, ab (-)
Inflammation
Unknown
18.0
8
Sterile during ab
Sterile, ab (+)
No
Inflammation
None
17.0
9
< 15 CFU*
Positive*
No
Inflammation
Yes
16.5
10
Sterile during ab
Sterile, ab (+)
No
Unknown
None
16.0
Identificationof the identicalmiroorganism;n . CFU = colonyformingunits.
d. =
not done;ab (+) = with antibiotictherapy;ab (-) = withoutantibiotictherapy;
to the missing or false d a t a m e n t i o n e d above, the low sensitivity of m i c r o b i a l c u l t u r e s h o u l d be m e n t i o n e d : T h e sensitivity of m u l t i p l e q u a l i t a t i v e a n d q u a n t i t a t i v e c u l t u r e s of c a t h e t e r tips is a b o u t 1 5 - 6 6 % a n d the predictive v a l u e a b o u t 1 5 - 3 7 % [4]. T h e positive predictive value for coagu l a s e - n e g a t i v e staphylococci of positive results in b l o o d c u l t u r e s is 4.1-26.4% [5]. I n g e n e r a l clinical practice n o t all c a t h e t e r tips are cul-
S 52
t u r e d a n d parallel s e m i q u a n t i t a t i v e b l o o d cultures, d r a w n c e n t r a l l y a n d peripherally, are n o t d o n e regularly. Thus, o n the basis of m i c r o b i o l o g i c a l c u l t u r e results alone, the incid e n c e of c a t h e t e r - r e l a t e d i n f e c t i o n s w o u l d a p p e a r to b e too low. This score is l e g i t i m a t e a n d m o s t s u i t a b l e for m u l t i c e n t e r studies, in which i n d i v i d u a l decisions of the local investigator s h o u l d b e avoided. T h e risk of o b s e r v e r bias m i g h t
Infection 27 (1999) Suppl. 1 9 MMV Medien & Medizin VerlagsGmbH M~inchen1999
S. Lugauer et al.: Catheter-Associated Infection Score System
be reduced by the chronological longitudinal documentation of selected raw data in different local investigation sites and a later central blinded analysis. This is possible, because all decisions and grading of diagnostic criteria can be made mathematically. The result can be reviewed any time and is free of subjective decisions. References National Center for Infections Diseases. Centers for Disease Control and Prevention: Guideline for prevention of intravascular device-related infections. Am. J. Infect Control. 24 (1996) 262-293.
2.
Gastmeier, P., Weist, K., Riiden, H.: Catheter-related primary bloodstream infections: epidemiology and methods of prevention. Infection 27 (Suppl. 1) (1999) S 1~6. 3. Martus, P., Gels, C., Lugauer, S., Biiswald, M., Guggenbichler, J.-P.: Clinical study of the Erlanger silver catheter - data management and biometry. Infection 26 (Suppl. 1) S. 61~58. 4. Meinhold, M., Kiihlwein, J,, Goldschmidt, 1t., Bach, A.: Vergleich yon verschiedenen Untersuchungsmethoden der mikrobiellen Besiedelung zentraler Katheter. 4. Workshop Katheterinfektionen, Mainz 1997. 5. Herwaldt, L. A., Geiss, M., Kao, C., Pfaller, M. A.: "lqaepositive predictive value of isolating coagulase negative staphylococci from blood cultures. Clin. Infect. Dis. 22 (1996) 14-20.
Infection 27 (1999) Suppl. 1 9 MMV Medien & Medizin VerlagsGmbH Mtinchen 1999
S 53