International guidelines: the need to standardize the management of candidaemia

Candidaemia is a condition associated with high mortality.[1] Several guidelines relating to its management were published between 2009 and 2013 [2–6] and as demonstrated in a recent critique, the guidelines differ significantly from each other.[7] These inconsistencies lead to clinical uncertainty. In this editorial, we elucidate the problem as it may occur in clinical practice. To determine the extent of variation among the five official guidelines [2–6] for the management of candidiasis, we collected data on all adult patients with proven candidaemia from 2011–2014. These examples from clinical practice show that attempts should be made to standardize the management of candidaemia. Our audit was carried out in NHS Ayrshire and Arran, Scotland, United Kingdom.

Fifty-six patients had candidaemia during the study period. The 30-d survival was 64.2%. Fifty-seven Candida species were isolated from 56 patients (one patient had mixed infection). These include C. albicans (n = 28), C. glabrata (n = 16), C. parapsilosis (n = 8), and one each of C. dubliniensis, C. guilliermondii, C. lusitaniae, C. kefyr, and C. krusei. Only four isolates were resistant to fluconazole (three isolates of C. glabrata and the C. krusei). Two patients with endocarditis had persistent candidaemia. Fifty-three out of the 56 patients were alive at the time of candidaemia detection. The results of this audit are therefore confined to these 53 patients all of whom received antifungal therapy.

Primary treatment was with an echinocandin (n = 37 [69.8%]), fluconazole (n = 14 [26.4%]), voriconazole (n = 1 [1.8%]), or a combination of lipid amphotericin and 5-flucytosine (n = 1 [1.8%]). Thus, adherence to the recommendations of the European Society for Clinical Microbiology and Infectious Diseases (ESCMID) and Brazilian Guidelines for the Management of Candidiasis (BZL) was modest (69.8%), as these guidelines prefer echinocandin therapy in all patients,[5,6] while adherence to German Speaking Mycological Society and the Paul-Ehrlich Society for Chemotherapy (GSMS-PES) guideline was very high (51 out of 53; 96.2%) as both fluconazole and echinocandin are offered as primary therapy in this guidance.[4] The Infectious Diseases Society of America (IDSA) and the Canadian Clinical Practice (CCP) guidelines recommend risk stratification to determine which patients should receive echinocandin. The IDSA recommends echinocandins in the subset of patients with diabetes, cancer, hemodynamic instability, advanced age, or previous azole use while the CCP guidelines consider hemodynamic instability as the predominant risk factor.[2,3] Forty-four patients had at least one of the five risk factors for initiation of therapy with echinocandins (Table 1, rows 5–9) (some had multiple risk factors). Of these 44 patients, 35 (77.2%) were treated with an echinocandin and were therefore compliant with the IDSA guidelines. Since all seven patients who were hemodynamically unstable were treated with an echinocandin, and none of the three patients with prior azole use (Table 1, row 5) were treated with fluconazole, we were fully compliant with the CCP guidelines. Thus, depending on which guideline is chosen as the audit standard, our compliance rate varied from 69.8% to 100%.

Table 1. Characteristics of patients and selected areas of management.

Row	Category	Number
1	Adult patients	56
2	Males/females	21/35
3	Mean age	64.7
4	Alive at the time of detection of candidaemia	53
5	Azole use in the previous 30 d	3
6	Cancer	15
7	Diabetes	7
8	Elderly	29
9	Haemodynamic instability	7
10	Any of the above risk factors in rows 5–9	44
11	None of the risk factors in rows 5–9	9
12	Initial therapy	Rows 13–16
13	Fluconazole	14
14	Echinocandin (caspofungin or micafungin)	37
15	Lipid amphotericin +5-FC	1
16	Voriconazole	1
17	Eligible for step-down fluconazole (sum of rows 14–16)	39
18	Eligible but not received step-down fluconazole	12
19	Patients who received step-down fluconazole	27
20	Patients who received step-down fluconazole when stable	26
21	Patients who received step-down fluconazole but not stable	1
22	Patients with CVC	29
23	CVC not removed	2
24	CVC removed (sum of rows 25–26)	27
25	CVC removed within 48 h	16
26	CVC removed after 48 h	11
27	Funduscopic examination	7
28	Echocardiogram (TTE or TOE)	15
29	TTE	14
30	TOE	3
31	Daily blood cultures	0

5-FC: 5-flucytosine; CVC: central venous catheter; TTE: transthoracic echocardiogram; TOE: transoesophageal echocardiogram.

Step-down therapy to fluconazole was instituted in 27 of the 39 (69.2%) patients who were commenced on agents other than fluconazole. The IDSA recommends stepping down from broad-spectrum therapy to narrow-spectrum oral fluconazole in clinically stable patients provided the isolates are susceptible.[2] All but one patient (i.e., 26 out of 27 patients who were de-escalated) were appropriately switched to azole therapy as per the IDSA guidance demonstrating a compliance of 96.2%. Applying the same dataset to the ESCMID guidelines, our compliance falls to 14.8% (4 out of 27 patients) because ESCMID recommends a minimum of 10 d of intravenous (IV) therapy for all patients with candidaemia.[5] However, ESCMID guidelines are unclear if the IV therapy exclusively refers to echinocandins or whether IV fluconazole is also considered acceptable as part of a two-step sequential step-down (i.e., IV echinocandin to IV fluconazole followed by a second switch to oral fluconazole after day 10). In the absence of a specific recommendation, we believe that the 10 d of IV therapy in the ESCMID guidance refers to echinocandin therapy. The majority (n = 23) of our patients were switched to fluconazole between days 3 and 9 which is broadly in line with the GSMS-PES guidance.[4] We were fully compliant with the CCP and the BZL guidelines which make no recommendations in relation to timing of step down.[3,6] Our compliance varied between 14.8 and 100% depending on the audit standard.

Central venous catheter (CVC) removal is advised by all guidelines. Twenty-nine patients had a CVC at the time of the candidaemia: 27 (93.1%) had the CVC removed. Current guidelines do not specify the time when this should occur. This is important because patients in whom CVC removal does not occur at or around treatment initiation will cause a bias towards higher compliance as a majority will have survived up until the time of CVC removal (which will happen eventually). An older British guideline from 2003 specifically advised removal within 48 h.[8]

Funduscopy was performed in seven (13.2%) patients only. Our compliance with this performance measure was poor with regards to the IDSA and ESCMID guidelines. However, the BZL guideline recommends funduscopy only if patients have visual symptoms and we complied fully with this guidance.[6] Historically, funduscopy has been used to diagnose disseminated candidiasis and to determine the risk/benefit ratio of commencing treatment with the nephrotoxic agent, amphotericin.[9] With the availability of potent and safer agents, the ocular complication rates of candidaemia are decreasing and so the value of routine funduscopy in all patients with candidaemia has been questioned.[10]

Transthoracic echocardiogram (TTE) was carried out in 14 (26.4%) patients. Only three (5.6%) patients had a transoesophageal echocardiogram (TOE), so we were poorly compliant with the ESCMID guidance which recommends TOE for all patients.[5] However, since we carried out TOE in the two patients who had persistent candidaemia, we were 100% compliant with the GSMS-PES guidance.[4] Interestingly, the endocarditis guideline from the British Society for Antimicrobial Chemotherapy (BSAC) recommends TTE and not TOE as the initial diagnostic modality.[11]

The length of treatment for our patients was variable. Sixteen patients received therapy for fewer than 14 d including the 12 patients who died during this period. For the patients who survived 14 d, compliance with GSMS-PES guideline that recommends a total of 14 days of therapy was 90.2%.[4] However, compliance with IDSA, ESCMID, and BZL guidelines, all of which advise 14 d treatment counting from the day of clearance of candidaemia, was much lower (48.8%) as the optimal duration of therapy was invariably prolonged because blood cultures subsequently reported negative were often taken a few days into therapy.[2,5,6] The recommendation around the length of treatment needs to be standardized.

Follow up blood cultures are recommended in patients with candidaemia varying from two drawings of cultures during the first week of therapy (days 3 and 5) [6] to alternate day blood cultures [2] or daily cultures.[5] None of our patients had follow up daily blood cultures.

Table 2 provides a comparison of our compliance against each of the five guidelines.

Table 2. Comparison of compliance data with guidelines.

Category	IDSA	CCP	GSMS-PES	ESCMID	BZL
Primary therapy	77.2%	100%	96.2%	69.8%	69.8%
Step-down therapy	96.2%	100%	85.1%	14.8%	100%
CVC removal	93.1%	93.1%	93.1%	93.1%	93.1%
Fundoscopy	13.2%	NA	13.2%	13.2%	100%
Echocardiogram	NA	NA	100%	5.6%	NA
Length of therapy	48.8%	NA	90.2%	48.8%	48.8%
Follow up blood cultures	0%	NA	NA	0%	0%

BZL: Brazilian guidelines for the management of candidiasis; CCP: Canadian clinical practice; CVC: central venous catheter; ESCMID: European Society of Clinical Microbiology and Infectious Diseases; GSMS-PES: German Speaking Mycological Society and the Paul-Ehrlich-Society for Chemotherapy; IDSA: Infectious Diseases Society of America.

Some aspects may affect the relevance of our work. Regional guidelines may not be worldwide applicable. However, all guidelines have used standard criteria for judging the strength and quality of evidence. Variation in the epidemiology of candidaemia in Germany, Europe as a whole, Canada, USA, and Brazil does not appear to influence the recommendations: greater proportion of C. parapsilosis in Brazil cannot explain the preferential echinocandin use as this species is considered less susceptible to echinocandins.[12] Thus, the extent to which regional epidemiology influences the recommendations of the individual guidelines is unclear. Moreover, guidelines from groups with global acclaim (e.g., ESCMID, IDSA) still significantly vary in their recommendations. The proportion of C. glabrata is high in both Europe and USA but the approach to echinocandin use in ESCMID and IDSA guideline is remarkably different which led to variation in our compliance figures.[2,5]

The increasingly complex nature of guidelines and the lack of uniformity contribute to uncertainties in clinical decision-making. Moreover, guidelines, although meant for different conditions, sometimes contradict each other in relation to a common theme e.g., the dichotomy between ESCMID guideline and the BSAC guidelines in relation to echocardiogram discussed above.[5,11] This audit demonstrates that recommendations in relation to candidaemia differ to such an extent that clinical decisions are variably compliant. We believe that international guidelines should be standardized for the management of candidaemia, particularly when they address common themes.

Disclosure statement

The other authors declare no relevant competing interests.