• Aucun résultat trouvé

Risk factors for treatment failure in scabies: a cohort study

N/A
N/A
Protected

Academic year: 2021

Partager "Risk factors for treatment failure in scabies: a cohort study"

Copied!
6
0
0

Texte intégral

(1)

MEDICAL DERMATOLOGY British Journal of Dermatology

Risk factors for treatment failure in scabies: a cohort study*

A. AussyiD,1E. Houivet,2V. Hebert,1H. Colas-Cailleux,1N. Laaengh,1C. Richard,1M. Ouvry,1C. Boulard,1 S. Leger,1

N. Litrowski,1J. Benichou,2P. Joly,1and the investigators from the Normandy Association of Medical Education in Dermatology

1

Department of Dermatology, Rouen University Hospital and INSERM U1234, University of Rouen Normandie, Rouen, France

2

Department of Biostatistics, Rouen University Hospital and INSERM U1219, University of Rouen Normandie, Rouen, France Linked Comment: Williams and Fuller. Br J Dermatol 2019; 180:710–711.

Correspondence

Audrey Aussy.

E-mail: Audrey.aussy@chu-rouen.fr

Accepted for publication

21 October 2018

Funding sources

None.

Conflicts of interest

None to declare.

*Plain language summary available online DOI 10.1111/bjd.17348

Summary

Background Treatment failure, which occurs in about one-third of cases, is considered as a major factor in the increasing incidence of scabies in developed countries. Objectives To identify predictors of treatment failure of scabies in ambulatory pop-ulations.

Methods This multicentre study compared the clinical characteristics and treatment modalities between a group of patients with scabies treated successfully and another group who were not cured 3 months after antiscabies treatment.

Results In total 210 patients with a diagnosis of scabies were included, comprising 98 patients in the treatment success group and 112 in the treatment failure group. The main risk factors for treatment failure were (i) the use of only one type of treatment, topical benzyl benzoate (BB) or oral ivermectin, vs. the combi-nation of both treatments [odds ratio (OR) 215, 95% confidence interval (CI) 122–377]; (ii) the use of a single intake (vs. two) of oral ivermectin (OR 102. 95% CI 449–232); (iii) intake of ivermectin during a meal vs. on an empty stomach (OR 431, 95% CI 189–984); (iv) absence of decontamination of fur-nishings (OR 872, 95% CI 350–218), in particular sofa and cushions (OR 590, 95% CI 234–149), mattresses (OR 416, 95% CI 135–128) or car seats (OR 657, 95% CI 327–132) and (v) absence of written documents explaining treatment modalities (OR 518, 95% CI 257–104). In multivariate analysis, treatment failure was mainly associated with (i) use of a single intake (vs. two) of ivermectin (OR 662, 95% CI 271–162); (ii) use of BB alone vs. two intakes of ivermectin (OR 351, 95% CI 155–795) and (iii) absence of decontamina-tion of furniture with acaricides (OR 581, 95% CI 196–167).

Conclusions Use of topical BB alone and a single intake (vs. two) of ivermectin are predictors of treatment failure.

What’s already known about this topic?

Scabies is a major public health problem worldwide.

Developing countries account for the majority of cases, but an increasing incidence of scabies has been reported in developed countries.

Treatment failure is considered a major factor in the increasing incidence of scabies in developed countries.

Risk factors for antiscabies treatment failure, such as dementia and bedridden sta-tus, have been identified in hospitalized populations and are not relevant in ambu-latory patients.

What does this study add?

This study, which was conducted in a large cohort of ambulatory patients, high-lights several risk factors.

(2)

A single intake of ivermectin is the major cause of treatment failure in scabies.

The combination of one topical application of benzyl benzoate and two intakes of

oral ivermectin results in the lowest rate of treatment failure.

Not disinfecting fomites increases the risk of treatment failure.

Scabies is a major public health problem worldwide, affect-ing approximately 130 million people in the world.1 Devel-oping countries account for the majority of cases, with prevalence rates between 13% and 46%.2 In these populations, overcrowding, poor hygiene and poverty are the main factors for reinfestation and treatment failure.2–4 The prevalence of scabies is estimated at between 2% and 4% in Europe and the U.S.A.3,4 Treatment failure, which occurs in about one-third of cases, is considered the major factor in the increasing incidence of scabies reported in the recent past in developed countries such as France and the U.K.5–8

Predictors of treatment failure in developed countries have been assessed exclusively in a series of hospitalized patients.9 These predictors – immune deficiency, dementia and bedrid-den status– are less relevant in ambulatory patients. As these latter represent the majority of cases of scabies, we conducted a study to identify predictors of treatment failure of scabies in ambulatory patients.

Patients and methods

Design and population

This observational study was conducted from June 2011 to December 2013 in the dermatology department of Rouen University Hospital and in 14 private dermatology practices in Normandy, a region in the north-west of France. Consecutive patients were included if they fulfilled the following criteria: (i) clinical diagnosis of scabies; (ii) agreement to fill in a stan-dardized form to collect data on the patient’s living condi-tions, family status and treatment modalities; and (iii) agreement to return for a follow-up visit 3 months after treat-ment initiation. This study was approved by the institutional review board of Rouen University Hospital, no. E2014-24.

Diagnosis of scabies was performed clinically. It was consid-ered definitive when mites or burrows were seen using der-matoscopy, and highly presumptive when the following signs of scabies were observed: presence of burrows associated with erythematous papules and/or pruritus with nocturnal predom-inance, and typical localizations such as the hands, flexural surface of the wrist, elbows, genitalia, umbilicus, nipples and buttocks.

Two groups of patients were defined at the time the study was designed. The treatment success group included patients who were referred for a first episode of scabies; success corre-sponds to complete healing defined by absence of both clinical

signs and itching, and no clinical features of scabies at the 3-month follow-up visit. The second group, the treatment failure group, included two subgroups of patients: (i) patients referred for a first episode of scabies and who had persistent clinical features of scabies at the 3-month follow-up visit, and (ii) patients who were referred for clinical symptoms of sca-bies despite an initial antiscasca-bies treatment prescribed in the previous 3 months, including topical application(s) of benzyl benzoate (BB) and/or oral ivermectin, which were the only two labelled antiscabies treatments at the time of the study in France. Antiscabies treatments were left to each investigator’s preference in this observational study.

Data collection

Data were recorded from standardized forms that patients had to fill in. The first set of items referred to patients and their living conditions: age, sex, rural or urban environment, place of residence, family status, sleeping habits, mode of childcare, socioeconomic status, relevant comorbidities and duration of symptoms before the diagnosis. A second set of items referred to treatment modalities for the first episode of scabies, includ-ing type of medication(s) used (BB and/or ivermectin alone or in association), number of repetitions of the treatment, use of acaricides to decontaminate the furniture (spraying A-PAR, containing Neo-Pynamin forte 43 g and Sumithrin 43 g on sofas, cushions, car seats and mattresses for 10–12 h) and the number of the patient’s relatives simultaneously treated. Finally, patients were asked to mention if they had received a prefilled electronic document explaining in detail the treat-ment modalities. Patients referred for a first episode of scabies filled in the form at the 3-month follow-up visit, whereas patients referred for scabies despite an initial antiscabies treat-ment prescribed in the previous 3 months filled in the form at the inclusion visit.

Statistical analysis

All of these data were compared between patients from the treatment failure group and patients from the treatment suc-cess group. The target sample size was 231 patients to detect an odds ratio (OR) of 25 for a prevalence of 30% of patients having treatment failure, with a 90% power and for a two-sided type I error of 5%.

The groups were compared using Student’s t-test for quanti-tative characteristics and Pearson’sv2-test or Fisher’s exact test as appropriate for categorical characteristics. All P-values were

(3)

two tailed and P-values < 005 were considered to be signifi-cant. Univariate analysis was performed using GraphPad Prism software (version 60; GraphPad Software, La Jolla, CA, U.S.A.). Variables with P < 020 in the univariate analysis were analysed by backward multivariate analysis in order to determine independent risk factors associated with treatment failure. Variables with > 15% missing data were excluded. Final model discrimination was assessed by estimation of the area of the receiver operator characteristic curve. Goodness of fit was assessed using the Hosmer and Lemeshow goodness-of-fit test. Multivariate analysis was performed with SAS soft-ware (version 93; SAS Institute Inc., Cary, NC, U.S.A.).

Results

Population of the study

We identified 231 patients of mean age 217  195 years with a highly presumptive (n = 89) or definitive (n = 142) diagnosis of scabies. No patient recruited in hospital refused to be included. The number of patients seen by office-based dermatologists who refused to be included has not been recorded. Among the 231 patients recruited, 154 were referred for a first episode of scabies, but 21 (9%) were sec-ondarily excluded because they did not attend the 3-month follow-up visit. Among the remaining 133 patients, complete healing (i.e. absence of clinical signs and absence of pruritus) was observed in 98 patients (737%), corresponding to the treatment success group. Thirty-five patients (263%) still pre-sented clinical features of scabies at the 3-month evaluation; these constitute the treatment failure group, together with the 77 patients referred for clinical symptoms of scabies despite an initial antiscabies treatment prescribed in the previous 3

months. This corresponds to a total of 112 patients in the treatment failure group. Finally, 210 patients (107 male, 103 female) were included (Fig. 1).

Univariate analysis of risk factors for treatment failure The mean age of patients (228  206 and 224  188 years; P= 089) and their distribution with a highly presump-tive (357% vs. 408%) or definipresump-tive diagnosis of scabies (643% vs. 592%, P = 058) were not different between the treatment failure and treatment success groups, respectively. The main risk factors for treatment failure related to character-istics of scabies and the patient’s living conditions were (i) duration of symptoms> 1 month before diagnosis [OR 397, 95% confidence interval (CI) 210–751], (ii) presence of other cases of scabies in the patient’s relatives (OR 182, 95% CI 105–316) and (iii) use of childcare regardless of mode (nursery or childminder’s home: OR 255, 95% CI 103– 632; babysitter in patient’s home: OR 212, 95% CI 107– 421) (Table S1; see Supporting Information).

We did not find any difference in the socioeconomic status (worker, self-employed, civil servant, employee, student, farmer, sales representative, executive manager, craftsperson, retired person) between the treatment failure and treatment success groups (data not shown). As 21 patients (9%) were lost to follow-up, we checked that the main demographics and clinical characteristics of scabies in these patients were not statistically different from those in the rest of the study popu-lation (data not shown).

The main risk factors for treatment failure related to ment modalities were (i) the use of only one type of treat-ment, topical BB or oral ivermectin, vs. the combination of both treatments (OR 215, 95% CI 122–377); (ii) the use of

(4)

a single intake of oral ivermectin vs. two intakes of ivermectin (OR 102, 95% CI 449–232); (iii) intake of ivermectin dur-ing a meal vs. on an empty stomach (OR 431, 95% CI 189– 984); (iv) treatment of only some vs. all of the patient’s rela-tives (OR 252, 95% CI 128–499); (v) absence of decontam-ination of furnishings (OR 872, 95% CI 350–218), in particular sofa and cushions (OR 590, 95% CI 234–149), mattresses (OR 416, 95% CI 135–128) or car seats (OR 657, 95% CI 327–132); (vi) absence of written documents (prefilled electronic document) explaining treatment modali-ties (OR 518, 95% CI 257–104) and (vii) the use of topical corticosteroids after treatment of scabies (OR 205, 95% CI 111–379) (Table S2; see Supporting Information). Con-versely, the use of one vs. two application of BB was not asso-ciated with a higher risk of relapse (OR 134, 95% CI 074– 244).

In the treatment failure group, most patients had been previously treated by a general practitioner (59 of 112 patients, 53%), whereas 23 (21%) had been previously trea-ted by private-practice-based dermatologists and 30 (27%) by hospital-based dermatologists. Among the 133 patients who were referred for a first episode of scabies and were followed up for 3 months, 45 were seen by a private-prac-tice-based dermatologist and 88 by a hospital-based derma-tologist. The rate of treatment success (98 patients, 736%) was not statistically different according to whether the patients were treated by a private-practice-based (31 of 45, 69%) or a hospital-based dermatologist (67 of 88, 76%; P= 021).

Multivariate analysis

The independent predictors of treatment failure were (i) use of BB alone vs. two intakes of ivermectin (regardless of whether ivermectin was associated or not with BB) (OR 351, 95% CI 155–795), (ii) use of a single intake of ivermectin vs. two (OR 662, 95% CI 271–162), (iii) absence of decon-tamination of furniture with an acaricide (OR 581, 95% CI 196–167), (iv) presence of other cases diagnosed in the patient’s relatives (OR 213, 95% CI 107–424) and (v) symptoms during> 1 month before diagnosis (OR 295, 95% CI 138–632) (Table 1).

Discussion

This study shows that the absence of a second intake of iver-mectin and the use of only one type of treatment (topical BB or oral ivermectin alone), in particular the use of BB alone, are the main predictors of treatment failure in scabies. Intake of iver-mectin during a meal also appears to be a predictor of treatment failure, which is in accordance with the manufacturer’s recom-mendations, which indicate taking ivermectin on an empty stomach. Interestingly, the number of applications of BB was not associated with an increased risk of treatment failure. Con-versely, items related to the patient’s socioeconomic conditions and potential financial difficulties in buying the antiscabies treat-ment were not associated with treattreat-ment failure.

Other strong predictors were related to the risk of recon-tamination, such as the absence of decontamination of furni-ture and presence of other cases of scabies in the patient’s relatives. On the other hand, the absence of treatment of all the patient’s relatives, use of childminders, absence of written information (prefilled electronic document) on treatment modalities and use of topical corticosteroids within the month following the treatment of scabies were significant only in univariate analysis.

The usefulness of decontamination of furniture is controver-sial in the literature and this practice is not universally adopted. However, this predictor was shown in both the uni-variate and multiuni-variate analyses in our study. The risk of treatment failure associated with the absence of written infor-mation on treatment modalities, which was seen in the uni-variate analysis, is likely due to patients’ misunderstanding of the chronology of administering oral treatments (and even more so for topical treatments) and decontaminating furnish-ings. The lack of standardized treatment instructions might theoretically have biased the results. However, most dermatol-ogists in France use the forms available on the websites of the National Health Insurance and the National Health Agency, which are in fact quite similar.10

It is likely that our findings correspond to those observed in clinical practice. The mean age of patients, 217 years, was in accordance with the highest proportion of scabies in the 10– 29-year-old population reported in previous epidemiological and therapeutic studies.11–15

Table 1 Independent predictors of treatment failure in scabies from multivariate analysis

Odds ratio estimate

(95% Wald confidence interval) P-value Delay between onset of pruritus and first scabies treatment> 1 month 295 (138–632) 00054 Other known case(s) of scabies in the patient’s entourage 213 (107–424) 00309 Absence of decontamination of furniture with acaricide 581 (196–167) 00014 One intake of oral ivermectin vs. two intakes (associated or not with BB) 662 (271–162) < 0001 Topical BB alone vs. two intakes of ivermectin (associated or not with BB) 351 (155–795) 00026 BB, benzyl benzoate. The area under curve of the receiver operating characteristic curve was 085, 95% confidence interval 080–091, and the Hosmer and Lemeshow goodness-of-fit test did not show any lack of fit (P= 014).

(5)

The previous predictors of treatment failure reported in the literature are immune deficiency, dementia and bedridden sta-tus, identified in an elderly population of hospitalized patients.9 This does not correspond to the outpatient popula-tion included in the present study. A study conducted in boarding schools in Cameroon also found that the presence of other cases of scabies in the patient’s close entourage was a risk factor of treatment failure.16

The beneficial effect of a second intake of ivermectin has already been described in different therapeutic trials.12–14The absence of ovicidal activity of ivermectin has been suggested to explain these results.12,13 Conversely, as reported by Ly et al., we found no evidence that the number of topical appli-cations of BB was a risk factor for treatment failure, regardless of whether the patient had also received ivermectin.15 This result has potential therapeutic implications as repeated appli-cations of BB frequently induce skin irritation.

This study was performed in a large population of consecu-tive outpatients recruited by a panel of private-practice-based and hospital-based dermatologists, making selection bias unli-kely. Additionally, the 3-month delay for assessment of treat-ment efficacy in the present study allowed us to take into account early relapses or recontaminations.

Various biases are possible in this study. In order to avoid selection bias, we aimed to recruit patients from both private-practice-based and hospital-based dermatologists. This is a major difference from previous reports, which assessed prognostic fac-tors exclusively in series of hospitalized patients. These studies identified immune deficiency, dementia and bedridden status as predictors of treatment failure. The mean age of patients in our study was 217  195 years, which indicates that the patients recruited by both hospital-based and private-practice-based der-matologists were ambulatory patients. Accordingly, we did not see significant differences in treatment outcomes according to whether patients were seen in hospital or private practice.

Whereas most patients referred to dermatologists for treat-ment failure had been previously treated by a general practi-tioner (53%), we did not see significant differences in the treatment outcomes of patients referred for a first episode of scabies, according to whether the patients were seen in hospi-tal or in private practice. Similarly, we did not see significant differences in the treatment outcomes according to whether the patients lived in rural or urban areas. The diagnosis of sca-bies was not confirmed by parasitological examination. How-ever, diagnosis of scabies is usually performed clinically in the majority of clinical trials.12–15,17

We cannot exclude that some patients in the treatment fail-ure group might have intentionally or involuntarily wrongly reported how their initial treatment was performed. In partic-ular, the fact that patients from the treatment failure group had to perform two applications of BB but may have applied only one could have biased the results, thus partly contribut-ing to the findcontribut-ing that two applications of BB were not better than one. However, even systematic checking by the patient’s pharmacist would not have confirmed whether the prescrip-tion had been respected or not.

Finally, we cannot exclude confusion bias. In particular, corti-costeroids may have been used in patients with more extensive or severe forms of scabies. Despite the fact that permethrin is now more commonly used than BB as a topical treatment, due to its better tolerance, previous studies showed that the efficacies of both drugs were very close.18 It is therefore likely that the same risk factors for failure of antiscabies treatment would have been identified with other antiscabies topical treatments.

This study highlights the fact that a single intake of iver-mectin, which is currently approved in many countries, is a major cause of treatment failure. This finding has potential implications as ivermectin is frequently prescribed instead of topical treatments in clinical practice, especially by general practitioners due to its higher convenience.17The combination of one application of topical BB and two intakes of oral iver-mectin was associated with the highest rate of success at the 3-month follow-up (86%) in the group of 133 patients not previously treated for scabies.

In summary, the association of two intakes of oral iver-mectin and one application of topical treatment improves the results of treatment and might be recommended to eradicate scabies in ambulatory populations in developed countries.

Acknowledgments

We acknowledge all of the investigators from the Normandy Association of Medical Education in Dermatology, in particular M. Ouvry, S. Bechu, N. Mion-Mouton, O. Lafaurie, A. Barrel, M. Pulluard, V. Hamel, M.X. Dore, E. Duflo, S. Lardans-Cassius, J.C. Rzeznik and S. Baricault. We are grateful to Nikki Sabourin-Gibbs, Rouen University Hospital, for her help in editing the manuscript.

References

1 World Health Organization. Investing to overcome the global impact of neglected tropical diseases: third WHO report on neglected tropical diseases. Available at: http://apps.who.int/iris/ bitstream/10665/152781/1/9789241564861_eng.pdf (last acce ssed 22 November 2018).

2 Fuller LC. Epidemiology of scabies. Curr Opin Infect Dis 2013;26:123–6. 3 Engelman D, Kiang K, Chosidow O et al. Toward the global

con-trol of human scabies: introducing the international alliance for the control of scabies. PLOS Negl Trop Dis 2013;7:e2167.

4 Romani L, Steer AC, Whitfeld MJ et al. Prevalence of scabies and impetigo worldwide: a systematic review. Lancet Infect Dis 2015; 15:960–7.

5 Owusu-Edusei K, Chesson HW, Gift TL. The economic burden of pediculosis pubis and scabies infections treated on an outpatient basis in the United States: evidence from private insurance claims data, 2001–2005. Sex Transm Dis 2009; 36:297–9.

6 Pannell RS, Fleming DM, Cross KW. The incidence of molluscum contagiosum, scabies and lichen planus. Epidemiol Infect 2005; 133:985–91.

7 Bitar D, Thiolet JM, Haeghebaert S et al. Increasing incidence of scabies in France, 1999–2010, and public health implications. Ann Dermatol Venereol 2012; 139:428–34.

8 Mounsey KE, McCarthy JS. Treatment and control of scabies. Curr Opin Infect Dis 2013; 26:133–9.

(6)

9 Makigami K, Ohtaki N, Ishii N et al. Risk factors for recurrence of scabies: a retrospective study of scabies patients in a long-term care hospital. J Dermatol 2011;38:874–9.

10 Haute Conseil de la Sante Publique. Recommandations relatives a la conduitea tenir devant un ou plusieurs cas de gale (Recommenda-tions on how to deal with one or more cases of scabies). Available at: https://www.hcsp.fr/Explore.cgi/Telecharger?NomFichier=hc spr20122209_conduitegale.pdf (last accessed 22 November 2018). 11 Lassa S, Campbell MJ, Benett CE. Epidemiology of scabies

preva-lence in the U.K. from general practice records. Br J Dermatol 2011; 164:1329–34.

12 Usha V, Gopalakrishnan Nair TV. A comparative study of oral iver-mectin and topical permethrin cream in the treatment of scabies. J Am Acad Dermatol 2000; 42:236–40.

13 Goldust M, Rezaee E, Raghifar R. Comparison of oral ivermectin versus crotamiton 10% cream in the treatment of scabies. Int J Der-matol 2014; 53:904–8.

14 Manjhi PK, Sinha RI, Kumar M et al. Comparative study of efficacy of oral ivermectin versus some topical antiscabies drugs in the treatment of scabies. J Clin Diagn Res 2014;8:HC01–04.

15 Ly F, Caumes E, Ndaw CA et al. Ivermectin versus benzyl benzoate applied once or twice to treat human scabies in Dakar, Senegal: a

randomized controlled trial. Bull World Health Organ 2009;87:424– 30.

16 Kouotou EA, Nansseu JR, Kouawa MK et al. Prevalence and drivers of human scabies among children and adolescents living and study-ing in Cameroonian boardstudy-ing schools. Parasit Vectors 2016;9:400. 17 Shimose L, Munoz-Price LS. Diagnosis, prevention, and treatment

of scabies. Curr Infect Dis Rep 2013;15:426–31.

18 Hlawa B. Treatment of scabies with permethrin versus lindane and benzyl benzoate. Acta Derm Venereol 1989;69:348–51.

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher’s website:

Table S1 Univariate comparison of the main clinical, demographic and environmental factors at baseline between the treatment failure and treatment success groups.

Table S2 Univariate comparison of treatment modalities and associated procedures between patients from the treatment failure group and treatment success groups.

Figure

Fig 1. Flowchart of the study.
Table 1 Independent predictors of treatment failure in scabies from multivariate analysis

Références

Documents relatifs

Preclinical Study of Single-Dose Moxidectin, a New Oral Treatment for Scabies: Efficacy, Safety, and Pharmacokinetics Compared to Two-Dose Ivermectin in a Porcine Model..

Sie haben eine Tochter oder einen Sohn mit Down-Syndrom (das Al- ter spielt keine Rolle) und stellen fest, dass es andere Kinder mit Down- Syndrom gibt, die sich im gleichen

Equilibrium calculations have been used successfully to compute the stable phase assemblages based on the solution composition as well as to model the stable phase assemblage

[r]

2 Faculté de pharmacie, Université de Montréal, Montréal 3 Institut universitaire de cardiologie et de pneumologie de Québec (IUCPQ), Québec 4 Faculté de pharmacie,

Based on such calibration performed on different waters used in drinking water production plants and the abrasion measurements done on the mem- branes that filter these waters, it

In this dataset each patient under study is assigned one of the studied treatment strategies, he/she is described by a set of initial features that can be nominal or numerical, and

Voici un exemple de paragraphe de développement dont les parties sont liées (le paragraphe porte sur les liens entre la nature et les émotions dans un extrait d’Atala