A Cross-Sectional Evaluation of the Usefulness of the Minor Features of Hanifin and Rajka Diagnostic Criteria for the Diagnosis of Atopic Dermatitis in the Pediatric Population

Abstract

Background:

Atopic dermatitis (AD) is a chronic, relapsing inflammatory skin disease. Hanifin and Rajka's criteria is the most common diagnostic criteria used for the clinical diagnosis of this condition. However, many find that it is too exhaustive to be used in routine practice, and the specificity of many of the minor criteria poses challenges, particularly from Asian countries with type III, IV, and V skin.

Aims and Objectives:

The aim of the study is to evaluate the effectiveness of the minor features of the Hanifin and Rajka criteria for AD in comparison to the UK working party's diagnostic criteria in pediatric populations of India.

Methodology:

A hospital-based cross-sectional study of 100 patients in the pediatric age group (3 months–12 years) with AD was conducted based on history, clinical, and laboratory evaluation. An age-matched control group of 100 pediatric patients who did not have a personal or family history of atopic dermatitis was included after obtaining informed consent to find out the prevalence of minor criteria among the control group.

Results:

Mean of the number of minor clinical criteria found positive in our study population in the infantile and toddler (below 2 years) and childhood groups (2–12 years) was (4.72 ± 1.75) and (5.67 ± 1.78), respectively. Early-onset of disease was the most consistent feature among the minor criteria found in 83% of patients, followed by xerosis (71%), hyperlinearity of palm (56%), pityriasis alba (54%), Denny Morgan fold (52%), elevated serum IgE (47%), perifollicular accentuation (37%), and tendency toward cutaneous infections (37%).

Conclusion:

We found that though some of the minor criteria are highly sensitive and specific to the diagnosis of AD (xerosis, ichthyosis, palmar hyperlinearity, tendency of cutaneous infections, Dennie–Morgan infraorbital fold, pityriasis alba, and perifollicular accentuation), some other criteria were either very rare or nonspecific for AD. We suggest that many of the minor criteria of Hanifin and Rajka may not have much significance for Indian patients and a multicentric nationwide study with a larger patient pool is required to create a trimmed and improved version of Hanifin and Rajka criteria.

Introduction

Atopic dermatitis (AD) is an itchy, chronic, or chronically relapsing inflammatory skin condition that often starts in early childhood (usually before 2 years of age), with the rash being characterized by erythema, itchy papules/papulovesicular (occasionally vesicles in infants), which may become excoriated and lichenified, and occurs in individuals who have a personal or family history of atopy.[1]

Currently, the diagnosis and assessment of the severity of AD are made mostly on clinical grounds in the absence of suitable diagnostic tests.[2] Hanifin and Rajka proposed diagnostic criteria for AD in 1980, which is still being used widely as the commonest diagnostic tool. The criteria, consisting of four major and 23 minor features (27 features in total), encompass a set of clinical symptoms and signs, aggravating or environmental factors, abnormal findings from invasive tests, ophthalmic findings, and personal or family history of atopic diseases.[3] In spite of the inherent problems, the Hanifin and Rajka criteria became a standard reference for clinical trials in AD because of their high sensitivity: 93% in the study by Williams et al.[4] and 96% in the study by De et al.[5]

However, several patients at a given moment may not present all the features necessary for a firm diagnosis of AD. While comprehensive and often utilized in clinical trials, such a large number of criteria are inconvenient for use in clinical practice. Some of the minor criteria have been noted to be poorly defined or nonspecific (such as pityriasis alba), while others, such as upper lip cheilitis and nipple eczema, are quite specific for AD but uncommon.[6,7,8,9]

Subsequently, the United Kingdom Working Group refined the Hanifin and Rajka criteria to improve their practical applicability, and these criteria included a minimum set of valid and reliable features that can be used in a hospital setting or a community setting.[10]

In one recently concluded study, it was found that Hanifin and Rajka's diagnostic criteria seem to be better in the diagnosis of AD than UK refinement criteria for young children.[11]

We conducted this study in a tertiary care center of eastern India to evaluate the frequency and significance of the minor criteria of Hanifin and Rajka criteria in comparison to the UK working party's diagnostic criteria in cases of atopic dermatitis of the pediatric age group in the Indian scenario.

Methodology

Patients

Children, aged 3 months–12 years, diagnosed with AD for the first time in the dermatology department of our tertiary care hospital from March 2018 to February 2020 were enrolled consecutively in the study as the patient group. The study was approved by the institute's ethical review committee. Informed consent was received from the parents/caregivers of the participants.

Clinical evaluation

All of the participants were evaluated by the same consultant dermatologists. The final diagnosis of AD was given by the consultants following the UK working party's diagnostic criteria.[10] The dermatologic diseases in the control group included various common skin disorders such as verruca, impetigo, seborrheic dermatitis, pityriasis versicolor, and chronic fungal skin infections. The medical and family histories were obtained from the caregivers and were recorded in the case record forms. A family history of atopy was defined as having at least one parent with asthma and/or allergic rhinitis diagnosed by the physician.

Evaluation of diagnostic criteria

Both patients and controls were investigated for the presence of minor criteria of Hanifin and Rajka. However, type 1 hypersensitivity by prick testing was not studied due to a lack of infrastructure and expertise. Additionally, “keratoconus” and “anterior subcapsular cataract” were evaluated with the help of ophthalmology colleagues. Total immunoglobulin E (IgE) was measured for all participants with the nephelometric method (Siemens Healthcare Diagnostics Products, Marburg, Germany). If IgE was above the cut-off levels according to age 18, it was defined as “high” IgE. “Early age of onset” was defined as the onset of symptoms of eczema before the age of 2 years.[12] Cutaneous infections were defined as the presence of at least two episodes of folliculitis or furunculosis or impetigo or diagnosed herpes simplex infection in the past 1 year. Nonspecific hand or foot dermatitis was defined as the presence of itchy lesions on one or both hands/feet with erythema and papules/vesicles or scaling, with or without oozing, crusting, fissures, or lichenification.[13] Keratosis pilaris was diagnosed as more than 20 follicular, keratotic papules involving at least posterolateral aspects of upper arms or thighs.[13,14,15] Similarly, palmar hyperlinearity was defined by the presence of more than 5 prominent lines longer than 1 cm running across the palm.[13] Anterior neck folds were defined as prominent horizontal skin crease(s) on the anterior aspect of the neck when the head is upright, and perifollicular accentuation was defined as dermatitis enhanced around hair follicles in two or more areas with a diameter of >5 cm.[13]

Clinical assessment of the severity of the disease was also determined with the help of the SCORAD and EASI index.

Control

The children referred to the dermatology clinic of our hospital with skin problems and who do not have a personal or family history of AD were included as the age-matched control group.

Results

We evaluated 100 consecutive patients of AD diagnosed with UK working party's diagnostic criteria which included 51 boys (51%) and 49 girls (49%) with no significant gender difference (P = 0.84). In our study, 39 patients (39%) were in the age group of below 2 years (infant and toddler AD) and 61 patients (61%) were in the age group of 2–12 years (childhood AD).

The mean age of onset in our study was 33.8 ± 28.0 months (range: 2 months–8 years). Thirty patients (30%) had urban habitus, and 70 (70%) had suburban or rural habitus. The duration of illness at presentation varied from 1 month to 96 months. Its mean duration was 16.5 ± 21.0 months. Mode of onset was found to be insidious in most cases (67%), while acute onset was found in 33% of cases.

In our study, the involvement of scalp, face [Figures 1 and 2], trunk [Figure 3], limbs, genitalia, and palmoplantar involvement was noted in 31%, 78%, 65%, 66%, 13%, and 20% of cases, respectively. Flexural involvement [Figure 4] was present in 48% of cases, while extensors were involved in 59% of cases; both flexural and extensor involvement were present in 39% of cases [Table 1].

Figure 1.

Infantile atopic dermatitis with involvement of face and cheeks

Figure 2.

Infantile atopic dermatitis with involvement of face and cheeks with xerosis

Figure 3.

Extensive involvement of back in an 8-year-old boy with extreme xerosis

Figure 4.

Involvement of flexures in a case of AD in a female child of 7 years of age. Note the presence of excoriated papules

Table 1.

Area of involvement of atopic dermatitis in different age groups

Area of Involvement	Infant and toddler AD (below 2 years)	Childhood AD (2-12 years)
Face	92.3%	65.6%
Trunk	51.2%	73.8%
Flexure	25.6%	68.9%
Extensor	38.5%	72.1%
Both Flexure and Extensure	18.0%	64.0%
Palmoplantar	18.0%	23.0%

Figure 5.

Palmar hyper linearity in a case of AD

Figure 6.

Tendency toward the development of cutaneous infections in a case of childhood AD with the presence of lip-lick cheilitis

Figure 7.

Perifollicular accentuation on the upper back in a case of AD

Personal history of allergic rhinitis (AR) was present in 22 patients, personal history of bronchial asthma (BA) was present in 17 patients, and 3 patients had both AR and BA. While considering the family history of atopy; 33 patients had AD, 18 patients had AR, and 8 patients had BA in their first-degree relatives. Both personal and family history of atopy was present in 12 patients.

Severity grading of all the patients of the study group was done according to the SCORAD score, and it was seen that the mean value of SCORAD was 23.5 ± 11.4. Mild [SCORAD <25], moderate [25–50], and severe [>50] AD were seen in 66%, 26%, and 8% cases, respectively. The mean value of EASI was 5.8 ± 5.2. Mild, moderate, and severe cases according to the EASI score were found in 71%, 26%, and 4% of cases, respectively.

More than three minor features of Hanifin and Rajka criteria were found positive in 94% of our study population. The mean of the number of minor clinical criteria found positive in our study population in the infantile and childhood group was (4.72 ± 1.75) and (5.67 ± 1.78), respectively.

In our study, we found that early onset of disease was the most consistent feature among the minor criteria found in 83% of patients, followed by xerosis (71%), hyperlinearity of palm (56%), pityriasis alba (54%), Denny Morgan infraorbital folds (52%), elevated serum IgE (47%), perifollicular accentuation (37%), and tendency toward cutaneous infections (37%) [Table 2].

Table 2.

Comparison of percentage of frequency of minor criteria found positive in our study and control population with significance

Minor features	Percentage positivity in the study population	Percentage positivity in control population	Significance (P)
Xerosis	71	15	<0.00001 (S)
Ichthyosis	19	2
Palmar hyperlinearity	56	6	<0.00001 (S)
Keratosis pilaris	2	1
Type 1 Hypersensitivity	Not done	Not done	NA
Elevated serum IgE	47	14	<0.00001 (S)
Early-onset of disease	83	13	NA
Cutaneous infections	37	2	<0.00001 (S)
Hand foot dermatitis	20	0	NA
Nipple eczema	6	0	NA
Cheilitis	9	3	0.136559 (NS)
Recurrent conjunctivitis	0	0	NA
Dennie–Morgan infraorbital fold	52	18	<0.00001 (S)
Keratoconus	0	0	NA
Cataract	0	0	NA
Orbital darkening	3	0	NA
Facial pallor	11	0
Facial erythema	4	0
Pityriasis alba	54	27	0.000101 (S)
Anterior neck folds	22	13	0.093959 (NS)
Itch when sweating	7	7	1 (NS)
Intolerance to wool and lipid	0	0	NA
Perifollicular accentuation	37	10	<0.00001 (S)
Food intolerance	0	0	NA
Course influenced by environmental/emotional factors	30	5	NA
White dermographism	0	0	NA

NA – not applicable, NS- not significant (P>0.05), S- significant (P<0.05), highly significant at P<0.01. P with Yates correction has been taken in those cases whose any values in either column are below 5

A high value of statistical significance (P < 0.05) was found when we compared the incidence of xerosis, ichthyosis, palmar hyperlinearity, tendency of cutaneous infections, Dennie–Morgan infraorbital folds, pityriasis alba, and perifollicular accentuation of the study population with that of control. An important finding of our study was that the incidence of cheilitis, anterior neck folds, and itching when sweating had no statistical significance when compared to the control population (P > 0.05).

In our study, we found a high value of statistical significance (P < 0.0001) when we compared the incidences of the course of AD being influenced by environmental factors in infantile and childhood age groups. However, no such statistical significance could be established on comparing the other minor features of Hanifin and Rajka criteria between the infantile and childhood age subgroups [Table 3].

Table 3.

Comparison of percentage of frequency of minor features found positive in our study subpopulations with status of significance

Minor features	Frequency of positivity in age subgroups of the study population		Status of significance (P)
	<2 years (n=39)	>2 years-12 years (n=61)
Xerosis	25	46	0.224205 (NS)
Ichthyosis	0	19	-
Palmar hyperlinearity [Figure 5]	22	34	0.94731(NS)
Keratosis pilaris	0	2	-
Type 1 Hypersensitivity	Not done	Not done	NA
Elevated serum IgE	21	26	0.272729 (NS)
Early-onset of disease	39	44	-
Cutaneous infection [Figure 6]	10	27	0.0599 (NS)
Hand foot dermatitis	7	13	0.681722 (NS)
Nipple eczema	2	4	0.890139 (NS)
Cheilitis	1	8	0.149873 (NS)
Recurrent conjunctivitis	0	0	NA
Dennie-Morgan infraorbital folds	18	34	0.349451 (NS)
Keratoconus	0	0	NA
Cataract	0	0	NA
Orbital darkening	0	3	-
Facial pallor	3	8	0.604701 (NS)
Facial erythema	3	1	0.3253 (NS)
Pityriasis alba	20	34	0.662803 (NS)
Anterior neck folds	6	16	0.20163 (NS)
Itch when sweating	1	6	0.322974 (NS)
Intolerance to wool and lipid	0	0	NA
Perifollicular accentuation [Figure 7]	15	22	0.808741(NS)
Food intolerance	0	0	NA
Course influenced by environmental/emotional factors	1	29	<0.00001(S)
White dermographism	0	0	NA

NA – not applicable, NS- not significant (P>0.05), S- significant (P<0.05), highly significant at P<0.01. P with Yates correction has been taken in those cases whose any values in either column is below 5

Another significant finding of our study was that 30% of patients had an exacerbation of lesions of AD with respect to seasonal factors—winter and summer exacerbation were noted in 20% and 10% patients, respectively.

Discussion

AD is a chronic inflammatory cutaneous disease causing great morbidity and psychological stress in both patients and their parents. The prevalence of clinical features and intensity of symptoms of AD may vary with genetic background, climate, geographical locations, food habits, socioeconomic status, availability of healthcare facilities, and many other factors.

The mean age of the patients in our study was 4.2 ± 3.4 years. A similar finding was observed in other studies on children aged 3 months–12 years.[12,13,14,15,16] Mean age at onset in our study was 33.8 ± 28 months, similar to the findings of another study from northern India.[5]

Our findings are similar to that of Dhar et al.,[13] that the face is predominantly involved in infants and toddlers and extensors are more commonly involved than flexors in both younger and older age group children. Among personal and family history of atopic disorders, AR was more prevalent than BA in our study, similar to other studies conducted in India and in the west.[17,18,19,20]

Various dermatoses found to be associated with patients with AD were scabies (14%), impetigo (6%), folliculitis (4%), seborrheic dermatitis (4%), chronic urticaria (4%), acute urticaria (3%), tinea capitis (1%), lichen nitidus (1%), and lichen striatus (1%). Among systemic diseases, bronchial asthma (17%) and valvular heart disease (1%) were found to be associated.

On review of the literature, a study by Dhar et al. (13)stressed the fact that a variety of dermatoses are associated and lead to the severity of AD through exacerbations and stress. A study of 550 patients of AD for a period of 2 years in India revealed fungal infections that are superficial mycoses, mostly recurrent and chronic, were present in 23.7% of cases of AD. Bacterial infections were present in 34.0% and response to treatment was good, but relapses were frequent. Viral infections were present in 15.5% and response to treatment was moderate, but recurrences were common. Pigmentary disturbances such as hypo and hyperpigmentation were present; 6.2% had post inflammatory depigmentation, 4.1% had widespread hyperpigmentation, and 8.2% had cutaneous amyloidosis. Drug reactions were present in 6.2% in the form of exaggeration of AD in three, fixed drug eruptions in two, and urticaria in one.[20]

Hanifin and Rajka's criteria consisted of four major criteria and 23 minor criteria. Three from each category are necessary for diagnosing AD. Despite being highly sensitive in diagnosing AD, the criteria were often criticized for the long list of minor criteria, some of which had been found to be rare and nonspecific by subsequent workers.[5,6,7,8,9]

More than three minor features of Hanifin and Rajka criteria were found positive in 94% of our study population. The mean of the number of minor clinical criteria found positive in our study population in the infantile and childhood group were (4.72 ± 1.75) and (5.67 ± 1.78), respectively. Earlier studies showed that four major features of Hanifin and Rajka criteria were found in 72% and six minor features were seen in over 96% of cases.[8]

The most common finding among minor criteria observed in our study was the early age of onset of the disease, which was seen in 83% of the cases, similar to previous studies (67.8%–82.9%).[9,12,16]

Xerosis (71%) and palmar hyperlinearity (56%) were other commonly observed criteria in our study. Similar higher values of palmar hyperlinearity (54%) and xerosis (80%) were also noted by Kanwar et al.[8] Criteria such as pityriasis alba and Dennie–Morgan infraorbital fold were also seen in higher frequency in our study, that is, 54% and 52%, respectively, like in earlier studies.[8] Agrawal et al.[19] had a lower value of occurrence of pityriasis alba, probably due to the higher mean age of the participants in their study.

Serum IgE was elevated in nearly half (47%) of our patients, similar to previous Indian studies.[16,19] Cutaneous infections were found in 37% of cases in our study, which was similar to a study by Nagaraja et al.[15] (36%) but much higher than that of Parthasarathy et al.[16] (2.9%).

The presence of perifollicular accentuation varied from 20% to 47.7% in various studies on AD, which matched our findings (37%).[7,16] Our patients with AD had a higher incidence of nonspecific hand/foot dermatitis (20%) compared to other studies (4.6%–12%).[15,16]

Anterior subcapsular cataract, keratoconus, and recurrent conjunctivitis were not observed among any of the children examined like in other studies.[15] Most studies noted higher winter aggravation of AD, like in our studies,[15,20] but one Indian study earlier reported commoner summer aggravation in Indian patients.[14] Our study showed a high incidence of ichthyosis (19%) like in western studies, and unlike most Indian studies.[9,15,16]

Most Indian studies reported a lower incidence of keratosis pilaris in pediatric populations like our study (2%), except one by Nagaraja et al.,[15] which showed a higher incidence of keratosis pilaris (33%). We had no patient with a history of intolerance to wool and lipid solvents, compared to other studies (28%–41%), possibly because of the local culture of wearing clothes and garments and climatic conditions prevalent in eastern India.[16] Again, nipple eczema was another criteria that had a very low incidence in our (6%) and other Indian studies.[15] Criteria such as food intolerance and white dermographism were not observed in any of the 100 cases in our study similar to other Indian studies.[7,15,16] For the presence of the anterior neckline, our study had higher values (22%) than most Indian studies (6%) but much lower values than western studies (75.2%).[9,15,16] Facial pallor was found in 11% of cases in our study. This was in corroboration with an Indian study done by Kanwar et al.,[8] where it was 14%.

We found that at least seven minor criteria were statistically very significant (P < 0.00001) in this cohort of patients with AD when compared to the age-matched control population; these include xerosis, ichthyosis, palmar hyperlinearity, tendency of cutaneous infections, Dennie–Morgan infraorbital fold, pityriasis alba, and perifollicular accentuation. We also found at least three minor criteria that had no significant association with the patients of AD, including cheilitis, anterior neck folds, and itching when sweating. Various earlier studies also emphasized that many minor criteria are nonspecific and some of the criteria are only rarely present in the patients of AD [Table 4]. Kanwar et al.[7,8] mentioned that six minor features, namely cheilitis, nipple eczema, perifollicular accentuation, white dermographism, recurrent conjunctivitis, and anterior neck folds, were not specific as they were encountered in patients with AD as often as in controls.

Table 4.

Comparison of minor criteria in our study population with some other Indian and overseas studies (percentage values)

Minor clinical criteria	Present study	Parthasarathy et al. (93)	Nagaraja et al. (12)	Dhar et al. (24)	Rudzki et al. (94)
Xerosis	71	67.2	76	80	85.2
Ichthyosis	19	11.5	4	-	-
Palmar hyperlinearity	56	67.8	23	54	-
Keratosis pilaris	2	4	33	46	-
Type 1 Hypersensitivity	-	-	-	-	-
Elevated serum IgE	47	52.9	-	-	65.7
Early-onset of disease	83	67.8	73	74	82.9
Cutaneous infections	37	2.9	36	62	65.3
Hand foot dermatitis	20	4.6	12	42	81.9
Nipple eczema	6	0	1	8	23.1
Cheilitis	9	7.5	3	6	56.9
Recurrent conjunctivitis	0	0	14	4	24.5
Dennie-Morgan infraorbital folds	52	71.8	63	82	78.1
Keratoconus	0	1.7 (suspect)	0	-	0
Cataract	0	0	0	-	14.3
Orbital darkening	3	6.9	12	32	53.4
Facial pallor	11	21.3	26	14	-
Pityriasis alba	54	57.5	34	78	20.8
Anterior neck folds	22	6.3	6	12	75.2
Itch when sweating	7	8	35	66	77.6
Intolerance to wool and lipid	0	2.3	41	28	71.3
Perifollicular accentuation	37	47.7	39	22	-
Food intolerance	0	0	0	-	74.4
Course influenced by environmental/emotional factors	30	8	44	26	68.4
White dermographism	0	0	40	12	84.2

We also found a higher influence of environmental factors in the course of AD in infants and toddlers than in older children (P < 0.001) However, no such statistical significance could be established by comparing the other minor features of Hanifin and Rajka's diagnostic criteria between the two age groups.

Conclusion

We found that though some of the minor criteria are highly sensitive and specific for the diagnosis of AD (xerosis, ichthyosis, palmar hyperlinearity, tendency of cutaneous infections, Dennie–Morgan infraorbital fold, pityriasis alba, and perifollicular accentuation), some other criteria were either very rare (recurrent conjunctivitis, keratoconus, and cataract) or nonspecific (cheilitis, anterior neck folds, and itching when sweating) for AD. Intolerance to wool or lipid and white dermatographism, though found in other Indian studies, were also not present in any of our patients. Our studies suggest that many of the minor criteria of Hanifin and Rajka may not have much significance for the diagnosis of AD in Indian patients. A multicentric nationwide study with a larger patient pool is required to create a trimmed and improved version of Hanifin and Rajka's diagnostic criteria for AD suitable for dermatologists and pediatricians of the Indian subcontinent.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.