Skip to main content

Table 4 Core scientific questions, rationale, vignette comparison and results

From: The classification of feeding and eating disorders in the ICD-11: results of a field study comparing proposed ICD-11 guidelines with existing ICD-10 guidelines

Core scientific question

Rationale

Vignette comparison

ICD-11 diagnosis

Results

1. Does the proposed addition of ARFID in the ICD-11 result in individuals with ARFID being more accurately distinguished from AN, and does the proposed addition of ARFID to ICD-11 reduce the number of individuals diagnosed with residual eating disorders (atypical, other specified, and unspecified)?

The proposal to include ARFID in ICD-11 raised the research question as to whether ARFID when it is associated with underweight status can be accurately distinguished from AN using proposed ICD-11 guidelines

Vignette 1C: AN vs Vignette 2A: ARFID

ICD-11 AN DX: 96.6% accuracy

ICD-10 AN DX: 93.7% accuracy

χ2 (1) = 1.38, p = .24

ICD-11 ARFID DX: 89.9% accuracy

ICD-10 ARFID DX*: 80.4% accuracy

χ2 (1) = 2.34, p = .13

ICD-11 ARFID DX vs AN DX:

χ2 (2) = 246.25, p < 0.001

Overall ICD-11 was equal to ICD-10, but ICD-10 “applicable” options are spread across four diagnoses*

G2 (4) = 7.32, p = .16

2. Can clinicians distinguish between ARFID and no eating pathology based on the proposed ICD-11 guidelines?

The addition of a “new” diagnosis always raises the question of whether the proposed disorder can be properly distinguished from no disorder. The core research question addressed by this comparison was whether clinicians could better distinguish between ARFID and cases that should not be assigned a diagnosis based on the proposed ICD-11 guidelines as compared to the range of eating disorder residual categories in ICD-10.

Vignette 2A: ARFID vs Vignette 2B: No DX

ICD-11 ARFID DX: 88.5% accuracy

ICD-10 ARFID DX*: 76.8% accuracy

χ2 (1) = 6.71, p < .01

ICD-11 No DX: 78.4% accuracy

ICD-10 No DX: 79.6% accuracy

χ2 (1) = 0.17, p = .68.

ICD-11 ARFID DX vs No DX

χ2 (2) = 190.00, p < 0.001

Overall ICD-11 Outperformed ICD-10

G2 (4) = 17.80, p < 0.01.

3. Some individuals present with atypical reasons for restricting eating, such as feeling uncomfortable when full. In such cases, can clinicians accurately distinguish between AN and ARFID based on the proposed ICD-11 guidelines?

The diagnostic guidelines for a new disorder must sufficiently differentiate it from other existing disorders. We tested whether the proposed inclusion of ARFID can be clearly distinguished from AN when the rationale for restricting intake is atypical (e.g., restricting eating because of stomach fullness or bloating.

Vignette 1C: AN vs Vignette 2C: ARFID

ICD-11 AN DX: 96.7% accuracy

ICD-10 AN DX: 97.0% accuracy

χ2 (1) = 0.02, p = .89

ICD-11 ARFID DX: 87.9% accuracy

ICD-10 ARFID DX**: 76.0% accuracy

χ2 (1) = 6.90, p < 0.01

ICD-11 AN DX vs ARFID DX:

χ2 (2) = 262.84, p < 0.001

Overall ICD-11 outperformed ICD-10

G2 (4) = 14.62, p < 0.01

4. ICD-11 proposes that a diagnosis of AN be retained until an individual has at least one year of stabilized sufficient weight gain and cessation of behaviors aimed at promoting weight loss. Does this rule improve diagnostic accuracy for AN over the course of recovery?

ICD-10 does not provide clear diagnostic guidance for recently weight restored individuals with AN, which results in substantial variability in whether an AN diagnosis is applied to cases that still exhibit significant symptoms but have gained weight to within a relevant weight reference (e.g., based on BMI or population quartile). ICD-11 proposes that the diagnosis of AN continue to be applied until the individual has achieved attitudinal and weight recovery for 1 year without the support of continuing care.

Vignette 1A: AN (with recovery not yet independently sustained for 1 year) vs Vignette 1B: no DX (AN with recovery independently sustained over 1 year)

ICD-11 AN DX: 84.6% accuracy for 1A

ICD-11 no DX: 38.4% accuracy for 1BĎ®

ICD-11 AN DX vs no DX

χ2 (2) = 46.82, p < .001

(No independent ICD-10 comparison because this rule is new to ICD-11)

Overall ICD-11 outperformed ICD-10

G2 (4)ϮϮ = 31.84, p < 0.0001

5. Is the proposal to include subjective binge eating in ICD-11 BN clinically useful and effective in reducing residual eating disorder diagnoses?

The ICD-11 recommendation to allow subjective binge eating to fulfill a part of the diagnostic requirements for both BN and BED was based on extant data suggesting that the threshold for an objective binge episode is arbitrary and clinical reports indicating that binge size does not predict distress or impairment.

Although intended to improve clinical utility, the ICD-11 inclusion of subjective binge eating could inadvertently make the diagnosis of BN or BED more difficult.

Vignette 3A: BN (with objective binge eating) vs Vignette 3B: BN (with subjective binge eating)

ICD-11 Objective BN DX: 84.3% accuracy

ICD-10 Objective BN DX: 82.2% accuracy

χ2 (1) = 0.23, p = .63

ICD-11 Subjective BN DX: 61.4% accuracy

ICD-10 Subjective BN DX***: 69.6% accuracy

χ2 (1) = 10.62, p < 0.001

ICD-11 objective BN DX vs subjective BN DX:

χ2 (1) = 20.25, p < 0.001

Clinicians were more accurate in diagnosing BN with objective binge eating

Overall ICD-11 outperformed ICD-10

G2 (2) = 10.90, p < 0.01.

6. Do the proposed guidelines for ICD-11 enable clinicians to accurately distinguish between BN and BED?

This question is prompted by the inclusion of the new category of BED in ICD-11.

Vignette 3A: BN vs Vignette 4A: BED

ICD-11 BN DX: 90.2% accuracy

ICD-10 BN DX: 83.3% accuracy

χ2 (2) = 8.73, p < 0.05

ICD-11 BED DX: 78.0% accuracy

ICD-10 BED “equivalent” DX: 70.7% accuracy

χ2 (2) = 2.05, p = .36

ICD-11 BN DX vs BED DX

χ2 (2) = 182.50, p < 0.001

ICD-10 BN DX vs BED DX****

Wide variability of DX since BED does not exist in ICD-10. χ2 (2) = 152.99, p < 0.001

Overall ICD-11 outperformed ICD-10

G2 (4) = 11.40, p < 0.05

7. Are the proposed ICD-11 guidelines for BED clinically useful in distinguishing BED from no disorder?

Similar to Question 2, given the addition of BED to the ICD-11, the question arises whether the proposed disorder of BED can be properly distinguished from no disorder.

Vignette 4A: BED vs Vignette 4B: No DX

ICD-11 BED DX: 82.4% accuracy

ICD-10 BED (equivalent) DX****: 72.5% accuracy

χ2 (2) = 6.71, p < 0.05

ICD-11 No DX: 80.3% accuracy

ICD-10 No Dx: 76.8% accuracy

χ2 (2) = 10.54, p < 0.01

ICD-11 BED vs No DX:

χ2 (2) = 203.40, p < 0.001

Overall ICD-11 outperformed ICD-10

G2 (4) = 18.24, p < 0.01

8. Do the proposed ICD-11 guidelines provide sufficient clinical guidelines to distinguish BN and BED regardless of weight status?

This comparison examined the impact of weight status on the diagnosis of BN and BED. According to both the proposed ICD-11 guidelines and the ICD-10 guidelines, weight status should not impact diagnosis of BN and BED. However, given that the majority of individuals who present with BED for clinical care are also overweight, this question is designed to assess whether clinicians are able to accurately distinguish between BN and BED regardless of weight status.

Vignette 3A: BN normal weight vs Vignettes 3C: BN with obesity

Vignette 4A: BED slightly overweight

vs Vignette 4C: BED with obesity

Vignettes 3C: BN with obesity vs Vignette 4C: BED with obesity

ICD-11 BN DX with obesity: 88.5% accurate

ICD-11 BN DX normal weight: 90.2% accurate

χ2 (2) = 3.25, p = .20

ICD-11 BED DX obese: 90.5% accurate

ICD-11 BED DX slightly overweight: 82.4% accurate

χ2 (2) = 8.90, p < 0.05

ICD-11 BED DX obese vs BN DX obese χ2 (2) = 213.70, p < 0.001

ICD-10 BED “equivalent” DX**** with obesity:

83.2% accurate

ICD-10 BED “equivalent” DX**** with slight

overweight status: 70.7% accurate

χ2 (2) = 7.64, p < 0.05

ICD-10 BN DX with obesity: 69.3% accurate

ICD-10 BN DX normal weight: 83.32% accurate

χ2 (2) = 8.18, p < 0.05

ICD-11 vs ICD-10 for BN DX with obesity

χ2 (2) = 17.43, p < 0.001

ICD-11 vs ICD-10 for BED DC with obesity /BED “equivalent” DX with obesity

χ2 (2) = 3.52, p = .17

Overall ICD-11 outperformed ICD-10 for obese individuals with either BN or BED

G2 (4) = 21.54, p < 0.001

  1. Note: AN anorexia nervosa, ARFID avoidant-restrictive food intake disorder, BN bulimia nervosa, BED binge eating disorder, DX diagnosis. *Accurate DX in ICD-10: atypical anorexia nervosa, feeding disorder of infancy or childhood, other eating disorder or eating disorder unspecified; **accurate DX in ICD-10: atypical anorexia nervosa, other eating disorder or eating disorder unspecified; ***accurate DX in ICD-10: atypical bulimia nervosa, other eating disorder or eating disorder unspecified; ****accurate DX in ICD-10: atypical BN, overeating associated with other psychological disturbances, other eating disorder, or eating disorder unspecified. ϮVignette 1B = 53.1% still diagnosed AN in ICD-11; ϮϮbecause of the different diagnostic labels included in ICD-11 versus ICD-10, it is not possible to have a direct, diagnosis by diagnosis comparison of the two systems, and disorders were grouped into anorexia nervosa, another diagnosis, or no diagnosis for this comparison