Ana səhifə

Supplementary Material for “Prevalence and clinical significance of dsm-5 Attenuated Psychosis Syndrome in adolescents and young adults in the general population: The Bern Epidemiological At-Risk (bear) study”


Yüklə 138.84 Kb.
tarix24.06.2016
ölçüsü138.84 Kb.
Supplementary Material for “Prevalence and clinical significance of DSM-5 Attenuated Psychosis Syndrome in adolescents and young adults in the general population: The Bern Epidemiological At-Risk (BEAR) study”



1. “Canton Bern” 2

2. “Sample” 3

2.1. Measures to improve response rate 3

2.2. Recruitment (supplementary Figure S1) 3

2.3. Representativeness of the sample (supplementary Table S1 and S2) 4

3. Help-seeking for APS (supplementary Table S3) 7

4. Effects of presence of attenuated psychotic symptoms (supplementary Table S4) 8

5. References Supplementary Material 9




1. “Canton Bern”


The Canton of Bern is one of 26 cantons (states) in Switzerland, a country in Europe. Switzerland is not part of the European Union, but finalized its official entry to the Schengen area in 2009, thereby abolishing passport and immigration controls for citizens of the other 25 European member states. According to Swiss Statistics (http://www.bfs.admin.ch/) for 2011, 30.3% of the 16–40-year-old permanent inhabitants of Switzerland have a non-Swiss nationality: altogether, 24.8% are from European countries, 29.9% of these are from German-, English- or French-speaking countries.
Swiss cantons have a permanent constitutional status and a high degree of independence. Each canton has its own constitution, and its own parliament, government, and courts. The cantons differ considerably, mainly in terms of population and geographical area. The Canton of Bern is the second largest canton in both surface area and population (approximately 1 million, including 306,017 citizens within the age range of 16–40 years of age). Canton Bern is a mixed urban-rural area; about two-thirds of the population lives in one of nine agglomerations. The largest city in Canton Bern is the city of Bern, with approximately 134,000 citizens and a population density of roughly 2,500 persons/km²1 (i.e., a population density about five times that of Dublin, comparable to that of Gothenburg, two-thirds that of Amsterdam, half that of Stockholm or The Hague, one-third that of Copenhagen, and one-quarter that of New York City).2
The immigration rate in Canton Bern is lower than that of the country as a whole, with only 19.5% of its 16–40-year-old permanent inhabitants having a non-Swiss nationality. According to 2011 Swiss Statistics, 12.5% of Bern’s permanent inhabitants are from European countries, with 31.1% of these from German-, English- or French-speaking countries. Ethnic characteristics are not provided by Swiss Statistics, but based on information on nationality (e.g., European, Russian, North American, and Australian), the rate of Caucasians can be assumed to exceed 95%. Only 4.3% of inhabitants come from South or Central America, the Caribbean, Africa, or Asia.
In our randomly drawn base sample provided by the KAIO (N=2,941), 4.5% immigrated from South or Central America including the Caribbean (0.6%), Africa (0.9%), South Asia (2.3%), or East and Southeast Asia (0.7%).

2. “Sample”

2.1. Measures to improve response rate


Following established epidemiological research procedures,3-5 first contact with eligible persons was established by an information letter, personally addressing each potential participant and, in the case of minors, their parents. Dispatching written information about study aims and conditions prior to first telephone contact had been shown to increase the response rate in telephone surveys.6,7 The letter also informed of non-report of findings to avoid violating the “right not to know” (that one might suffer from a mental illness).8 The “right not to know” relates to one’s decision to refuse to acquire knowledge about oneself that is discussed as part of a person’s autonomy. While those who argue for the right or even obligation to know argue that knowledge is necessary for autonomy, their opponents argue that autonomy is precisely what grounds the right not to know, and that one’s autonomy might be protected by one’s refusal to obtain information because one is ensured an “open future.”9
An increase in response rate has also been repeatedly shown when incentives are used.6,10-12 Incentives extend a strong positive effect on response rates of persons with a lower educational background12 and/or little interest in the study topic.13 Since persons of lower educational background are more likely to refuse study participation, an incentive should counteract this selection bias.12 The amount of an effective incentive depends on the data being assessed as well as interview duration and should be higher for long interviews and sensitive data.10 Accordingly, the present study, which collected sensitive data and had an average interview duration of approximately 45 min, warranted a high incentive. While awarding an incentive after study participation is generally regarded as preferable to doing so beforehand,10 an incentive can either be offered to each participant directly or via participation in a lottery. As the present study included a large sample requiring a high incentive, a lottery with a main prize of 2,000 CHF was chosen over individual awards for cost-effectiveness reasons. This incentive was announced in the information letter.

2.2. Recruitment (supplementary Figure S1)


Of the initial sample (N=2,941), 1,966 were eligible (supplementary Figure S1). Using the definitions of the American Association for Public Opinion Research,14 the response rate was 66.4% (supplementary Figure 1). Among the 1,341 participants, interviewers discontinued 108 (8.1%) interviews prematurely, mainly due to insufficient language skills. Only 4 (0.2%) participants refused to continue the interviews. Thus, 1,229 (91.7%) interviews were completed (supplementary Figure S1). The 625 remaining eligible persons were either unable to participate in the interview (n=3) or refused (n=622) (supplementary Figure S1), citing lack of interest in the topic (52.9%), lack of time (44.5%), expectation of the assessment of too intimate data (15.3%), and excessive interview length (13.2%) as reasons for refusal; 38.9% provided no reason.


Supplementary Figure S1: Survey outcome rates of the BEAR study according to the definitions of the American Association for Public Opinion Research14

Note: The contact rate was related to both eligible cases and an estimate of eligible subjects among cases of unknown eligibility. The cooperation rate included partial interviews.



2.3. Representativeness of the sample (supplementary Table S1 and S2)


The eligible sample differed from the 16–40-year-old general population of Canton Bern in terms of age distribution, as available telephone landlines were most common among individuals aged 36–40 years (supplementary Table S1). The fewest landlines were identified for those least likely to live with either their parents still or their own families already (i.e., those aged 26–30 years; supplementary Table S1). While eligibility bias for these two age groups were of strong-to-moderate effect, eligibility bias for the other 3 age groups were only small (supplementary Table S1).
Those who completed participation did not differ from those who did not in terms of age, gender, or ratio of Swiss nationality (supplementary Table S2). Completed interviewees only differed significantly from the discontinued interviewees in nationality, since naturally, more interviews with non-Swiss than Swiss persons were discontinued for language-related reasons (supplementary Table S2). As a carryover effect of the age-related eligibility bias, the 1,229 interviewees differed equally from the 16–40-year-old general population of Canton Bern in terms of age distribution, and, related to the high number of 36–40-year-olds, were more likely to be married or cohabitating (Table 1). Other response effects in age-group distribution were only small (Table 1). Furthermore, because of the language-related exclusion criterion, interviewees more frequently had Swiss nationality (Table 1), though this response effect was even less than small (w=0.08). There was no difference in gender distribution between interviewees and the general population. Thus, as no response bias was detectable beyond the age-related inclusion bias in the 26–30- and 36–40-year-olds, the interviewees were regarded as sufficiently representative of their age group within the general population. This partial age-group bias would only be relevant if age effects were detected (see the main text for results). Further demographic data are provided in Table 1 of the main text.


Supplementary Table S1: Comparison of the eligible sample with the Canton Bern general population according to the Swiss Statistics Web site, maintained by the Federal Statistical Office (http://www.bfs.admin.ch)




Canton Bern statistic

(year of latest statistic)



Eligible sample

(N=1966)


One-dimensional ²-test

Age ranges (%)


mean age: 28.6 years (2009)

mean age: 30.8 years








16-20 years

18.2

15.1 (w=0.17)a







21-25 years

18.6

15.6 (w=0.16)







26-30 years

19.9

12.7 (w=0.36)

2(4)=186.251, p<0.001




31-35 years

20.5

22.4 (w=0.09)







36-40 years

22.8

34.2 (w=0.50)




Gender; % male

48.8 (2011)

47.5

2(1)=1.269, p=0.26, w=0.03

Nationality; % Swiss

86.3 (2011)

89.1

2(1)=13.001, p<0.001, w=0.03

a Effect size was the effect size index, w, which measures the discrepancy between paired portions over the cells with w=0.1 equalling a small effect, w=0.3 a medium effect, and w=0.5 a large effect.


Supplementary Table S2. Comparison of the interviewees versus refusers and complete versus partial interviews




Interviewees

(N=1341)


Refusers

(N=622)


²-test

Complete interview

(N=1229)


Partial interview

(N=112)


²-test, U

Age; mean (SD), range

30.73 (7.28), 16-40

30.96 (7.46), 16-40

U=406748.500 p=0.38, r=0.020a

30.66 (7.32), 16.6-41.4

31.75 (7.46), 17.1-40.4

U=63393.00 p=0.17, r=0.038a

Gender; % male

47.6

47.4

2(1)=0.004, p=0.95, V=0.001b

47.7

46.4

2(1)=0.065, p=0.80, V=0.007b

Nationality; % Swiss

87.1

90.0

2(1)=3.601, p=0.06, V=0.043b

93.1

56.2

2(1)=154.839, p<0.001, V=0.340b

a Effect size was Rosenthal’s r with r=0.1 equalling a small effect, r=0.3 a medium effect, and r=0.5 a large effect

b Effect size was Cramer’s V with V=0.1 equalling a small effect, V=0.3 a medium effect, and V=0.5 a large effect

3. Help-seeking for APS (supplementary Table S3)





Supplementary Table S3. Spontaneously named main reasons for any help-seeking in persons reporting any APS (n=61; 104 help-seeking contacts) and no APS (n=223; 322 help-seeking contacts), multiple answers possible

Main reasons for help-seeking

any APS

no APS

One-dimensional ²-test

(df=1)b






n

%a

n

%a

Depressive mood

46

44.2

102

31.7

2.058

Partnership and familial problems

23

22.1

91

28.3

0.763

Anxiety

22

21.2

51

15.8

0.788

Agitation

16

15.4

47

14.6

0.021

Withdrawal

13

12.5

19

5.9

2.367

Loss of energy

13

12.5

37

11.5

0.042

Tension

12

11.5

44

13.7

0.192

Other conduct abnormalities

8

7.7

6

1.9

3.504

Self-confidence

6

5.8

21

6.5

0.040

Sense of guilt

6

5.8

7

2.2

1.620

Irritability

6

5.8

14

4.3

0.223

Other affective alterations

6

5.8

10

3.1

0.819

Loss of appetite or sleep

5

4.8

28

8.7

1.127

Headache

4

3.8

7

2.2

0.427

Self-injurious behaviour

4

3.8

4

1.2

1.352

Hypersensitivity

3

2.9

8

2.5

0.030

Drug abuse

2

1.9

11

3.4

2.650

Memory problems

2

1.9

2

0.6

0.676

Antisocial behaviour

2

1.9

6

1.9

0.000

Expansive / manic mood

1

1.0

0

0

1.000

Loss of libido

0

0

4

1.2

1.200

Obsessive ideas

0

0

2

0.6

0.600

Alcohol abuse

0

0

7

2.2

2.200

Cognitive disturbances

0

0

3

0.9

0.900

Perceptual disturbances

0

0

0

0

-

Disorganized communication

0

0

0

0

-

Unusual thought content

0

0

0

0

-

Ideas of reference

0

0

0

0

-

Persecutory ideas / suspiciousness

0

0

0

0

-

Perceptual abnormalities

0

0

0

0

-

Odd behaviour

0

0

0

0

-

Hallucinations

0

0

0

0

-

Delusions

0

0

0

0

-

a related to number of help-seeking contacts

b critical ²(1) for p<0.05 is >3.842

4. Effects of presence of attenuated psychotic symptoms (supplementary Table S4)



Supplementary Table S4. Effects of presence of attenuated psychotic symptoms (n=159), Attenuated Psychosis Syndrome (n=4) and Attenuated Psychosis Syndrome-Revised (n=32) on sociodemographic and clinical variables




Attenuated psychotic symptoms
(lifetime) (1)

Attenuated Psychosis Syndrome (2)

Attenuated Psychosis Syndrome-Revised (3)

Effect sizes
1 vs. 2d

Effect sizes

2 vs. 3d






Yes

(n=159)


No

(n=1,070)



²(df), U,
effect size

Yes

(n=4)


No

(n=1,225)



²(df), U,
effect size

Yes

(n=32)


No

(n=1,197)



²(df), U,
effect size

²(1)

²(1)

Age; median (range)

32
(16-40)

32
(16-40)

U=83420.5 p=0.775, r=0.008a

21
(19-39)

32
(16-40)

U=1700.0 p=0.290, r=0.030a

36
(19-40)

32
(16-40)

U=13232.0 p=0.003, r=0.085a

0.008

0.665

Age ranges; n (%)





































16-20 years

27 (17.1)

171 (16.0)

2(4)=4.151, p=0.386, V=0.058b

2 (50.0)

196 (16.0)

2(4)=4.564, p=0.335, V=0.061

3 (9.4)

195 (16.3)

2(4)=10.630, p=0.031, V=0.093

1.274

2.630




21-25 years

24 (15.2)

168 (15.7)

1 (25.0)

191 (15.6)

3 (9.4)

189 (15.8)




26-30 years

20 (12.7)

132 (12.3)

0 (0)

152 (12.4)

1 (3.1)

151 (12.6)




31-35 years

27 (17.0)

247 (23.1)

0 (0)

273 (22.3)

6 (18.8)

267 (22.3)




36-40 years

61 (38.6)

353 (33.0)

1 (25.0)

413 (33.7)

19 (59.4)

395 (33.0)

Gender (male); n (%)

68 (43.0)

518 (48.4)

2(1)=1.567, p=0.211, V=0.036b

2 (50%)

584 (47.7)

2(1)=0.009, p=0.926, V=0.003b

18 (56.2)

568 (47.5)

2(1)=0.967, p=0.325, V=0.028b

2.792

2.016

Nationality (Swiss); n (%)

146 (92.4)

998 (93.2)

2(1)=130, p=0.719, V=0.010b

3
(75.0)

1,141 (93.1)

2(1)=2.039, p=0.153, V=0.041b

29
(90.6)

1,115 (93.1)

2(1)=0.309, p=0.579, V=0.016b

1.884

1.097

Family history of psychosis (1st-degree); n (%)

5 (5.6)

36 (3.2)

2(1)=0.118, p=0.731, V=0.010b

1(25)

40 (3.3)

2(1)=5.786, p=0.016, V=0.069b

1 (3.1)

40 (3.4)

2(1)=0.006, p=0.940, V=0.002b

4.406c

6.323c

Current SOFAS score; median (range)

81.5
(40-93)

89
(40-99)

U=39387.0 p<0.001, r=0.314a

81
(40-91)

88
(40-99)

U=643.0 p=0.010, r=0.074a

80
(40-90)

88
(40-99)

U=6893.5 p<0.001, r=0.179a

14.845c

4.358c

Highest-past-year SOFAS score; median (range)

83
(40-92)

89
(40-99)

U=40455.0 p<0.001, r=0.307a

80
(65-85)

88
(40-99)

U=624.5 p=0.009, r=0.075a

80
(40-90)

88
(40-99)

U=7327.5 p=0<0.001, r=0.173a

14.090c

3.873c

Difference between highest and current SOFAS scores; median (range)

0
(0-28)

0
(0-30)

U=79172.5 p=0.001, r=0.092a

0

(0-10)


0
(0-30)

U=1948.0 p=0.082, r=0.050a

0
(0-28)

0
(0-30)

U=18410.5 p=0.358, r=0.026a

1.242

0.758

SOFAS≤70; n (%)

28 (17.7)

21 (2.0)

2(1)=89.346, p<0.001, V=0.270b

1 (25.0)

48 (3.9)

2(1)=4.629, p=0.031, V=0.061b

12 (37.5)

37 (3.1)

2(1)=96.396, p<0.001, V=0.280b

13.197c

14.065c

a Effect size was Rosenthal’s r with r=0.1 indicating a small effect, r=0.3 a medium effect, and r=0.5 a large effect

b Effect size was Cramer’s V with V=0.1 indicating a small effect, V=0.3 a medium effect, and V=0.5 a large effect

c For p<0.05, the critical value for ²(1) is >3.842

d Effect sizes were compared with a one-dimensional ²-test with df=1

5. References Supplementary Material


  1. Schweizerische Eidgenossenschaft, Department für Umwelt, Verkehr, Energie und Kommunikation, Bundesamt für Raumentwicklung (ARE) http://www.are.admin.ch/themen/agglomeration/ Accessed January 15, 2013.

  2. Wikipedia contributors. List of urban areas by population. Wikipedia, The Free Encyclopedia. May 28, 2013, at 18:10 UTC. https://en.wikipedia.org/wiki/List_of_urban_areas_by_population Accessed June 6, 2013.

  3. Prince M. Cross-sectional surveys. In: Prince M et al. (eds.) Practical Psychiatric Epidemiology. Oxford, New York: Oxford University Press; 2003:111-130.

  4. Camburn D, Lavrakas PJ, Battaglia MP, Masses J, Wright RA. Using advance respondent letters in random-digit-dialing telephone surveys. http://www.amstat.org/sections/srms/Proceedings/y1995.html, p.969-974, Accessed September 10, 2013.

  5. Koponen P, Aromaa A. Survey design and methodology in national health interview and health examination surveys. http://ec.europa.eu/health/ph_projects/2000/monitoring/fp_
    monitoring_2000_annexe14_04_en.pdf. Accessed September 15, 2012.

  6. Edwards P, Roberts I, Clarke M. Increasing response rates to postal questionnaires: Systematic review. BMJ 2002;324:1-9.

  7. Kessler RC, Little RJ, Groves RM. Advances in strategies for minimizing and adjusting for survey nonresponse. Epidemiol Rev 1995;17:192-204.

  8. Adorno R. The right not to know: an autonomy based approach. J Med Ethics 2004;30:435-440.

  9. Bortolotti L, Widdows H. The right not to know: the case of psychiatric disorders. J Med Ethics 2011;37:673-676.

  10. Gelman A, Stevens M. Chan V. Regression Modeling and Meta-Analysis for Decision Making: A Cost-Benefit Analysis of Incentives in Telephone Surveys. J Business Econom Stat 2003;21:213-225.

  11. Groves RM. Survey Errors and Survey Costs. New York, NY: Wiley; 1989.

  12. Guyll M, Spoth R, Redmond C. The effects of incentives and research requirements on participation rates for a community-based preventive intervention research study. J Prim Prevention 2003;24:25-41.

  13. Roberts PJ, Roberts C, Sibbald B, Torgerson D. The effect of direct payment or a lottery on questionnaire response rates: a randomised controlled trial. J Epidemiol Community Health 2000;54:71-72.

  14. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 6th ed. Lenexa, KS: American Association for Public Opinion Research; 2009. http://www.aapor.org/Standard_Definitions/1481.htm. Accessed December 10, 2012.

--



Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©atelim.com 2016
rəhbərliyinə müraciət