Sufficiently Important Difference for Common Cold: Severity Reduction

Bruce Barrett; Brian Harahan; David Brown; Zhengjun Zhang; Roger Brown

doi:10.1370/afm.698

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Download figure
Open in new tab
Download powerpoint
Download figure
Open in new tab
Download powerpoint
Download figure
Open in new tab
Download powerpoint
Figure 1.
Sufficiently important difference (SID) by treatment scenario.
Note: These data represent 2 samples: 91 people interviewed in-person (entry and exit) and 162 people interviewed by telephone.

View popup

Table 1.

Participant Characteristics in Prospective Study (Intake and Exit Interviews)

Characteristics	Value
GED = general educational diploma; BA/S = bachelor’s degree; MA = master’s degree.
* 45 telephone interview participants were missing all demographic data.
† Self-reported household income.
Telephone interviews, No.
Calls	983
Interviews*	162
Prospective in-person study, No.
Enrolled	91
Exited	91
Age, years
Range	18–74
Mean (SD)	34.8 (13.3)
Sex, No. (%)
Women	139 (67.8)
Men	66 (32.3)
Ethnicity, %
White	68.4
African American	7.1
Hispanic	2.0
Asian	0.8
Other	3.6
Not reported	18.2
Annual income bracket, No. (%)
<$15,000	83 (32.8)
$15,000 – <$25,000^†	33 (13.0)
$25,000 – <$50,000	36 (14.2)
$50,000 – <$75,000	31 (12.2)
$75,000 – <$100,000	15 (5.9)
≥ $100,000	6 (2.4)
No response	49 (19.4)
Highest education level, %
Some high school	3.2
High school degree/GED	28.5
Some college	18.6
BA/S	20.6
MA or greater	10.3
Other or no response	19.0
Tobacco use, %
Nonsmoker	60.9
Current	7.1
Past	14.2
No response	17.8

View popup

Table 2.

Sufficiently Important Difference (SID) Severity Benefit by Interview Method and Treatment Scenario Type

Characteristic	Intake	Exit	Telephone	Intake + Telephone
CI = confidence interval.
Note: data distributions tended to skew to the negative (6 of 8 had mean values of ≤ 50%).
No. of interviews	91	91	162	253
Jackson score, mean (SD)	9.7 (3.6)	1.5 (2.6)	9.6 (3.7)	9.6 (3.7)
Unadjusted mean SID
Overall (SD)	0.40 (0.20)	0.41 (0.16)	0.40 (0.17)	0.40 (0.18)
95% CI	0.38–0.44	0.36–0.44	0.37–0.43	0.38–0.42
Accepting, regardless of severity benefit, % (SD)	15.9 (0.37)	17.6 (0.38)	15.3 (0.36)	15.5 (0.36)
Refusing, regardless of severity benefit, % (SD)	12.6 (0.33)	15.4 (0.36)	14.5 (0.35)	13.8 (0.35)
	Vitamin	Herbal Extract	Lozenge	Antiviral
Unadjusted mean SID
Sick groups, No. (SD)	0.25 (0.22)	0.32 (0.24)	0.47 (0.26)	0.57 (0.30)
95% CI	0.23–0.27	0.30–0.34	0.43–0.51	0.53–0.61
Accepting, regardless of severity benefit, % (SD)	30.4 (46.1)	15.4 (36.2)	4.7 (21.3)	4.7 (21.3)
Refusing, regardless of severity benefit, % (SD)	3.1 (17.5)	7.1 (25.8)	14.6 (35.4)	37.2 (48.4)
Group and Jackson score, adjusted SID
Sick groups (SD)	0.23 (0.22)	0.31 (0.38)	0.49 (0.45)	0.63 (0.50)
95% CI	0.19–0.27	0.27–0.35	0.43–0.55	0.56–0.69

View popup

Table 3.

Potential Explanatory Factors of Sufficiently Important Difference (SID) by Treatment Scenario

	Vitamin		Herbal		Lozenge		Antiviral
Factor	β (SD)	P Value	β (SD)	P Value	β (SD)	P Value	β (SD)	P Value
Note: Results of multivariate regression analysis of associations of SID with covariates (potential explanatory factors).
Intercept	0.27 (0.13)	.032	0.49 (0.13)	<.001	0.23 (0.16)	.137	0.45 (0.18)	.012
Female	−0.05 (0.03)	.198	0.04 (0.04)	.292	0.09 (0.04)	.037	0.08 (0.05)	.112
Smoking status
Never	—	—	—	—	—	—	—	—
Cigarettes, >5/d	0.00 (0.05)	.925	−0.01 (0.05)	.813	0.08 (0.06)	.227	−0.11 (0.07)	.106
Cigarettes, <5/d	−0.02 (0.05)	.758	−0.10 (0.06)	.092	0.04 (0.07)	.586	−0.08 (0.08)	.312
Past	−0.03 (0.06)	.634	0.02 (0.06)	.814	0.05 (0.08)	.494	−0.06 (0.09)	.476
Ethnicity
White	—	—	—	—	—	—	—	—
Other	−0.08 (0.08)	.318	−0.10 (0.08)	.244	0.01 (0.10)	.895	−0.01 (0.11)	.943
African American	−0.08 (0.09)	.365	−0.04 (0.09)	.704	0.03 (0.11)	.822	0.02 (0.13)	.877
Hispanic	−0.03 (0.13)	.803	−0.16 (0.14)	.238	0.11 (0.17)	.499	0.05 (0.19)	.783
Asian	−0.14 (0.17)	.415	−0.10 (0.18)	.569	0.01 (0.21)	.961	0.24 (0.24)	.318
Education level
<High school degree	—	—	—	—	—	—	—	—
High school degree	−0.22 (0.08)	.007	−0.08 (0.09)	.348	−0.02 (0.10)	.882	0.08 (0.11)	.504
Some college	−0.15 (0.08)	.071	−0.07 (0.09)	.417	−0.03 (0.11)	.809	0.15 (0.12)	.217
College degree	−0.21 (0.06)	.018	−0.13 (0.09)	.175	−0.08 (0.11)	.471	0.04 (0.13)	.749
≥ Graduate school	−0.15 (0.09)	.121	−0.10 (0.10)	.301	−0.06 (0.12)	.593	0.09 (0.13)	.482
Annual income
<$15,000	—	—	—	—	—	—	—	—
$15,000–<$25,000	0.10 (0.10)	.339	−0.07 (0.11)	.521	0.04 (0.13)	.789	0.15 (0.15)	.299
$25,000–<$50,000	0.14 (0.10)	.160	−0.03 (0.10)	.772	0.10 (0.12)	.431	0.18 (0.14)	.192
$50,000–<$75,000	0.12 (0.10)	.198	−0.11 (0.10)	.274	0.04 (0.12)	.726	0.20 (0.14)	.153
$75,000–<$100,000	0.08 (0.10)	.440	−0.14 (0.11)	.204	0.13 (0.13)	.303	0.13 (0.14)	.356
≥ $100,000	0.13 (0.10)	.191	−0.15 (0.10)	.158	0.08 (0.12)	.545	0.16 (0.14)	.257
Jackson score	0.00 (0.00)	.679	0.00 (0.00)	.779	0.00 (0.01)	.850	0.00 (0.01)	.565

Additional Files

Figures
Tables

The Article in Brief

Sufficiently Important Difference for Common Cold: Severity Reduction

Bruce Barrett, MD, PhD , and colleagues

Background In this study, people with colds were interviewed about 4 cold treatments. They were asked about the amount of benefit they would want in exchange for the costs and possible side effects of the treatments. The study set out to determine the sufficiently important difference (SID) in these treatments, that is, the smallest benefit that a treatment would require in order to justify its costs and risks.

What This Study Found Of the 4 treatments, a $0.05 vitamin C pill with few or no side effects required the least benefit to justify treatment, followed by a $0.50 dose of an herbal extract with no side effects (other than bad taste), a $0.20 lozenge, which could taste bad and cause nausea, and finally a $2 prescription-only pill with unknown side effects. The study finds that, on average, people want the severity of a cold to be reduced by 25% to 57% to justify the costs and risks of cold treatments.

Implications

This study sheds light on patients� health values, which play a role in medical decision making and the design of research studies.
The authors suggest that the concept of sufficiently important difference be tested for other medical conditions.