Sage Journals: Discover world-class research

Abstract

Objective: To present Australian normative data on the ten-item Kessler Psychological Distress Scale (K10).

Method: Analysis of cross-sectional data from the 2007 Australian National Survey of Mental Health and Wellbeing, a nationally representative household survey of 8841 adults. Mean K10 scores and K10 scores at selected percentiles of the K10 score distribution are presented by sex, age, the presence of mental disorders and the presence of physical conditions. Stratum-specific likelihood ratios were computed to help clinicians and researchers calculate predicted probabilities of mental disorder given scores on the K10.

Results: Scores on the K10 were generally higher in women compared to men, in people with a mental disorder compared to without a mental disorder and in people with affective disorders compared to people with substance use disorders. The SSLRs were informative in ruling in a diagnosis of mental disorder, particularly at the high or very high end of the psychological distress spectrum.

Conclusions: These data may be helpful for clinicians and researchers alike in understanding the likelihood of mental disorder in a given individual or sample.

Keywords

Kessler Psychological Distress Scale mental disorder norms

Measurement of psychological distress is vital to monitor the mental health of both individuals and populations at a given point in time and in determining the extent and determinants of trends in mental health over time. The Kessler Psychological Distress Scale (K10) was developed by Kessler and others [1–3] for use in the annual US National Health Interview Survey to measure the non-specific aspects of psychological distress (see http://www.hcp.med.harvard.edu/ncs/k6_scales.php for more detail). The range of research that has subsequently utilized the K10 and its shortened version, the K6, is broad and includes studies in both general population and clinical samples for both health monitoring and intervention outcome measurement. For example, the K10 has been used in samples of injecting drug users [4], individuals with epilepsy [5], individuals with HIV [6] as well as in indigenous [7] and rural communities [8–14]. This measure has also been translated into eleven different languages [15,16] and used as a screening tool for mental disorders in a variety of situations [7,17–24]. More recently, studies have emerged establishing the reliability and validity of the K10 across a number of diverse settings [4,6,15,16,25–28].

In Australia the K10 has also been used in state-based population surveys including the annual New South Wales Health Survey [29], the 2000 South Australian Health and Wellbeing Survey [30], the 2000 Health and Wellbeing Survey in Western Australia [31] and the annual Victorian Population Health Survey [32]. Nationally, the K10 has been included in the 2001, 2004–2005 and 2007–2008 National Health Surveys [33–35] and the 2001 Indigenous Health Survey (Australian Bureau of Statistics, 2002). The K10 currently plays an important role in the assessment of the outcomes of services provided to people with mental disorders and as such has been included as a data item in the Mental Health National Outcomes and Casemix Collection [36]. The K10 is also integral to the evaluation of the Better Outcomes in Mental Health Care programme, an initiative funded by the Australian Government designed to address the observed under-treatment of people with mental disorders [37].

Of particular interest to the current study is the use of the K10 in the 1997 and 2007 Australian National Surveys of Mental Health and Wellbeing (NSMHWB). Data from the 1997 NSMHWB on the correspondence between K10 scores and Composite International Diagnostic Interview-defined ICD-10 diagnoses of anxiety, affective and substance use disorders have been previously presented [38]. While these data provided the first published Australian norms for the K10 and have been a valuable source of information, a number of factors point to the need for updated normative data. Firstly, the results reported in Andrews and Slade [38] divide the K10 into equally spaced five-point bands that do not consistently overlap with now fairly well-established and validated cut-points on the K10. Secondly, the results presented in Andrews and Slade [38] do not report data split by sex and age, two important demographic factors known to be associated with psychological distress [39]. Thirdly, Andrews and Slade [38] report the prevalence of mental disorders separately for each K10 band yet provide no estimates of relative risk or likelihood of having a mental disorder given a score in a particular K10 band. While information of this kind from the 1997 NSMHWB has been published elsewhere [40] the utility of the K10 as a screening instrument would benefit from having all this information in a single source publication. Underlying these methodological issues is the assumption that the most effective research endeavours, government policy initiatives and clinical service models should be informed and shaped by the most up-to-date data available.

Given the widespread use of the K10 in Australia and the shortcomings of previous reports of normative data on the K10 the aim of the current paper is to report the most current normative data on the K10 from the 2007 Australian NSMHWB. Specifically, we aim to:

Report normative data on the K10 both in the total Australian population and split by sex and age.

Report the prevalence of mental disorders in those scoring in previously established and validated K10 strata.

Report the likelihood of having a mental disorder given a score in each K10 strata and demonstrate how this information can aid in diagnostic decisions.

Method

Sample and measures

The 2007 Australian National Survey of Mental Health and Wellbeing (NSMHWB) is a nationally representative epidemiological survey of the Australian adult population. Data were collected via face-to-face interviews by the Australian Bureau of Statistics (ABS). The response rate was 60%, the total sample containing 8841 individuals aged 16 and above. A more detailed discussion of the sampling frame and procedures is available elsewhere [41,42].

The survey employed a multistage sampling design and utilized a modification of Version 3.0 (also known as the World Mental Health version) of the World Health Organization Composite International Diagnostic Interview (CIDI) [43]. The CIDI assesses all diagnostic criteria for the major lifetime mental disorders including affective disorders (major depressive disorder, bipolar disorders, dysthymia), anxiety disorders (agoraphobia, generalized anxiety disorder, obsessive–compulsive disorder, panic disorder, post-traumatic stress disorder, social phobia) and substance use disorders (alcohol and drug harmful use and dependence). Disorders were assessed using the definitions and criteria of the Diagnostic and statistical manual of mental disorders, fourth edition (DSM-IV) [44] and the International statistical classification of mental and behavioral disorders, tenth revision (ICD-10) [45]. Mental disorder information is presented with respect to ICD-10. Equivalent data for DSM-IV yield almost identical results and are available from the corresponding author. Organic exclusion rules were imposed in making all diagnoses. Estimates of the prevalence of 12-month mental disorders were derived from a lifetime diagnosis plus answers to disorder-specific recency questions (e.g. ‘how recently did you have symptoms of …?’). The CIDI has been shown in a number of clinical calibration studies to have good concordance with standardized clinical assessments [46,47].

Any 12-month physical condition was defined as the presence of at least one of the six physical conditions recognized as Australian National Health Priority Areas (NHPAs). These were diabetes, asthma, coronary heart disease (CHD), stroke, cancer and arthritis. Respondents were first asked if they had ever been told by a doctor or nurse that they had one of these chronic physical conditions. If they said yes they were then asked whether they had that condition or received treatment for that condition in the twelve months prior to the survey. Answers to these latter questions were used to define any 12-month physical condition.

The K10 assessed indicators of psychological distress in the 30 days prior to interview using a five-value response option for each item ranging from one (none of the time) through to five (all of the time). The maximum score, indicating high psychological distress is 50. Absence of psychological distress is indicated by a minimum score of 10. Two respondents were not included in all analyses due to answers of ‘don't know’ to the K10 items. The K6 is made up of a subset of six items from the K10 with total scores ranging from six to 30. The exact wording of the 2007 NSMHWB version of the K10 (and K6) questions and the coding and scoring of the response categories can be found in the Appendix. (available online at http://informahealthcare.com/doi/abs/10.3109/00048674.2010.543653).

Statistical analysis

SAS 9.2 [48] and STATA 10.0 [49] were used to analyse the data. Data were weighted to match the composition of the Australian general population. Standard errors and 95% confidence intervals were calculated using jack-knife repeated replication [50] which provides unbiased estimates of the sampling error arising from complex sample selection procedures. Preliminary analyses revealed statistically significant differences in mean K10 scores across sex and age bands. In a multivariate linear regression analysis both sex (F(1, 59) = 39.07, p < 0.0001) and age band (F(6, 54) = 23.07, p < 0.0001) were significantly associated with scores on the K10. Therefore, mean K10 scores, associated standard errors and percentiles of the K10 score distribution were calculated by sex and age bands. Mean K10 scores, associated standard errors and percentiles of the K10 score distribution were also calculated for groups defined by the presence of 12-month mental disorder groups (any affective disorder, any anxiety disorder, any substance use disorder and any mental disorder), the number of mental disorder groups, the presence versus absence of any physical condition and groups defined by the comorbidity between mental disorders and physical conditions. Given the negative or right-skewed nature of the K10 score distribution the percentiles chosen for presentation were the 10th, 25th, 50th (median), 75th, 90th, 95th and 99th percentiles. The K10 scores were then categorized into four strata according to standard cut-points representing low (10–15), moderate (16–21), high (22–30) and very high (31–50) psychological distress [51]. The probability of having 12-month mental disorder groups was calculated within each K10 stratum. This was carried out in the total sample as well as separately for men and women.

Stratum-specific likelihood ratios (SSLRs) [52] were then calculated. These indicate by how much the odds of having the target disorder increase or decrease given a score in a particular stratum of the K10 and are calculated by dividing the percentage of people with mental disorder in a given stratum by the percentage of people without mental disorder in that stratum. SSLRs greater than 1 indicate that a score in a particular stratum is associated with a greater likelihood of mental disorder while scores less than 1 indicate that a score is associated with a lesser likelihood of mental disorder. In general, likelihood ratios above 10 provide sufficient evidence of the presence of the disease of interest while those below 0.1 provide sufficient evidence of its absence [53]. SSLRs are arguably more informative than traditional screening test indices such as sensitivity and specificity for two reasons. Firstly, SSLRs are more flexible in that they can be adapted to tests with more than two strata, while sensitivity and specificity can only be calculated from dichotomous tests. Secondly, unlike sensitivity and specificity, SSLRs can be used to predict the probability of mental disorder in an individual, information which is vital for clinicians and researchers. As described in [40] SSLRs and their 95% confidence intervals were calculated using spreadsheets available from http://www.ebpcenter.com.

Results

The mean and standard error of the K10 score as well as the chosen percentiles of the K10 score distribution are presented by sex and age in Table 1. In the total sample the mean K10 score was 14.5 (SE = 0.1) and the median K10 score was 13 with an interquartile range (IQR) (the range of scores between the 25th and the 75th percentiles of the K10 score distribution) of 11–16. In the total sample and across all age groups women had higher mean K10 scores than men. Mean K10 scores and percentiles of the K10 score distribution are presented by sex and major classes of ICD-10 mental disorder as well as by physical condition and groups defined by mental/physical comorbidity in Table 2. As expected, mean K10 scores were higher in people with a mental disorder (19.1, SE = 0.2) compared to people without a mental disorder (13.3, SE = 0.1). In general, people with an affective disorder had the highest mean K10 score (23.2, SE = 0.5), followed by people with an anxiety disorder (19.8, SE = 0.3), then people with a substance use disorder (18.0, SE = 0.5). Scores on the K10 increased as people experienced more classes of 12-month mental disorder; however, this increase was more pronounced for men than it was for women. In general, K10 scores were more strongly related to the presence of mental disorders than they were to the presence of physical conditions.

Tabel 1.

Mean K10 score and K10 score at selected percentiles of the K10 distribution by sex and age

		Total sample							Men								Women
		K10 score at selected percentiles of the K10 distribution							K10 score at selected percentiles of the K10 distribution								K10 score at selected percentiles of the K10 distribution
	Mean (SE) K10 score	10	25	50	75	90	95	99	Mean (SE) K10 score	10	25	50	75	90	95	99	Mean (SE) K10 score	10	25	50	75	90	95	99
Total sample Age (years)	14.5 (0.1)	10	11	13	16	21	25	35	14.0 (0.1)	10	11	12	16	19	24	34	15.0 (0.1)	10	11	13	17	22	26	37
Age (years)
16–24	14.9 (0.2)	10	11	14	17	21	25	35	14.0 (0.2)	10	11	13	16	19	23	31	15.8 (0.3)	11	12	14	18	23	28	37
25–34	14.8 (0.2)	10	11	14	17	22	25	32	14.5 (0.2)	10	11	13	16	20	25	32	15.1 (0.2)	10	12	14	17	23	25	31
35–44	15.0 (0.2)	10	11	13	17	22	27	37	14.5 (0.3)	10	11	13	16	21	25	41	15.5 (0.3)	10	11	14	17	24	28	37
45–54	14.8 (0.3)	10	11	13	16	22	26	39	14.2 (0.3)	10	10	13	15	20	25	36	15.3 (0.4)	10	11	13	17	23	26	39
55–64	13.9 (0.2)	10	10	12	16	20	25	34	13.5 (0.2)	10	10	12	15	19	24	31	14.3 (0.3)	10	11	13	16	22	25	35
65–74	13.1 (0.2)	10	10	11	14	18	24	35	12.7 (0.3)	10	10	11	14	17	20	36	13.5 (0.2)	10	10	12	14	19	25	35
75–85	13.1 (0.2)	10	10	12	15	17	21	28	12.6 (0.3)	10	10	11	14	17	20	25	13.6 (0.2)	10	11	12	16	18	22	30

Table 2.

Mean K10 score and K10 score at selected percentiles of the K10 distribution by 12-month ICD-10 mental disorder classes, 12-month physical conditions and physical/mental comorbidity groups

		Total sample							Men								Women
		K10 score at selected percentiles of the K10 distribution							K10 score at selected percentiles of the K10 distribution								K10 score at selected percentiles of the K10 distribution
Disorder status	Mean (SE) K10 score	10	25	50	75	90	95	99	Mean (SE) K10 score	10	25	50	75	90	95	99	Mean (SE) K10 score	10	25	50	75	90	95	99
Mental disorder classes
Affective disorder	23.2 (0.5)	14	17	22	29	35	39	42	22.9 (0.8)	14	16	22	28	35	40	43	23.4 (0.5)	14	17	23	29	35	39	42
Anxiety disorder	19.8 (0.3)	12	14	18	24	31	36	41	19.4 (0.6)	11	13	18	24	30	35	41	20.0 (0.4)	12	14	18	24	31	36	42
Substance use disorder	18.0 (0.5)	11	13	16	21	28	32	41	17.5 (0.7)	11	12	15	21	26	32	41	19.0 (0.8)	13	14	17	23	29	35	39
Any mental disorder	19.1 (0.2)	12	13	17	23	30	34	41	18.2 (0.4)	11	13	16	22	28	33	41	19.7 (0.3)	12	14	18	24	31	35	41
No mental disorder	13.3 (0.1)	10	11	12	15	18	21	27	13.1 (0.1)	10	10	12	14	17	20	27	13.6 (0.1)	10	11	12	15	19	22	27
Mental disorder comorbidity
One mental disorder	17.4 (0.2)	11	13	16	20	26	30	37	16.3 (0.4)	11	12	15	19	24	28	34	18.2 (0.3)	12	13	16	21	27	31	39
Two mental disorders	23.5 (0.7)	14	17	23	29	35	39	42	22.2 (0.9)	14	16	21	26	34	37	42	24.6 (0.8)	15	18	24	31	37	39	42
Three mental disorders	26.9 (2.0)	16	20	25	32	41	41	48	28.9 (2.9)	17	24	26	38	41	41	48	23.9 (1.9)	15	17	23	30	36	38	42
Physical conditions
Any physical condition	15.0 (0.1)	10	11	13	17	22	26	36	14.4 (0.2)	10	11	13	16	21	25	34	15.4 (0.2)	10	11	14	18	24	27	39
No physical condition	14.2 (0.1)	10	11	13	16	20	25	35	13.8 (0.1)	10	11	12	15	19	23	33	14.7 (0.1)	10	11	13	16	22	26	36
Physical/Mental comorbidity
Neither	13.2 (0.1)	10	11	12	15	18	21	27	12.9 (0.1)	10	10	12	14	17	19	27	13.5 (0.1)	10	11	12	15	18	21	27
Physical condition only	13.6 (0.1)	10	11	12	15	19	22	28	13.3 (0.1)	10	10	12	15	18	21	28	13.8 (0.2)	10	11	13	16	19	22	28
Mental disorder only	18.5 (0.3)	11	13	17	22	29	35	41	17.7 (0.5)	11	12	16	21	27	33	41	19.2 (0.4)	12	14	17	23	31	35	39
Both	20.1 (0.4)	12	15	18	25	30	34	43	19.5 (0.6)	12	13	18	24	30	34	43	20.5 (0.5)	12	15	19	25	30	35	44

SE, standard error.

The prevalence of mental disorder by sex and K10 strata is shown in Table 3. Across all mental disorder groups the prevalence increased as the K10 score increased, with close to four out of five respondents (79.6%, SE = 3.6) in the very high strata experiencing a mental disorder in the twelve months prior to the survey. This information should be interpreted in light of the total sample distribution of K10 scores. Just over 70% of the total sample had a K10 score in the low range. A further 20% had a score in the moderate range, 7% in the high range and 3% in the very high range.

Table 3.

Prevalence of 12-month ICD-10 mental disorders and associated stratum specific likelihood ratios amongst individuals scoring from low to very high on the K10

	Low K10 score (10–15) (n = 6281; 71.1% (SE = 0.7) of total sample)		Moderate K10 score (16–21) (n = 1709; 19.5% (SE = 0.7) of total sample)		High K10 score (22–29) (n = 622; 6.9% (SE = 0.4) of total sample)				Very high K10 score (30–50) (n = 227; 2.6% (SE = 0.2) of total sample)
Mental disorder status	% (SE)	SSLR (95%CI)	% (SE)	SSLR (95%CI)	% (SE)	SSLR (95%CI)	% (SE)	SSLR (95%CI)	% (SE)
Total sample
Affective disorder	1.7 (0.2)	0.3 (0.2–0.3)	9.1 (0.8)	1.5 (1.3–1.7)	27.2 (2.7)	5.6 (4.8–6.6)	53.9 (4.6)	17.6 (13.8–22.6)	6.2 (0.3)
Anxiety disorder	7.1 (0.5)	0.5 (0.4–0.5)	23.2 (1.3)	1.8 (1.6–2.0)	44.7 (2.7)	4.8 (4.2–5.6)	67.2 (3.7)	12.2 (9.3–16.0)	14.4 (0.5)
Substance use disorder	3.5 (0.3)	0.7 (0.6–0.7)	7.0 (0.8)	1.4 (1.2–1.6)	13.1 (1.7)	2.8 (2.2–3.5)	14.6 (3.6)	3.2 (2.2–4.5)	5.1 (0.3)
Any mental disorder	10.9 (0.5)	0.5 (0.5–0.5)	32.0 (1.3)	1.9 (1.7–2.1)	57.1 (2.6)	5.3 (4.6–6.2)	79.6 (3.6)	15.6 (11.4–21.4)	20.0 (0.5)
Men
Affective disorder	1.4 (0.4)	0.3 (0.2–0.3)	8.6 (1.6)	1.7 (1.4–2.1)	28.6 (5.0)	7.2 (5.6–9.2)	59.1 (6.7)	25.8 (17.3–38.5)	5.3 (0.5)
Anxiety disorder	5.3 (0.6)	0.5 (0.4–0.5)	19.0 (2.1)	1.9 (1.7–2.2)	43.9 (5.0)	6.5 (5.1–8.2)	58.5 (7.6)	11.6 (7.7–17.5)	10.8 (0.7)
Substance use disorder	4.8 (0.6)	0.7 (0.6–0.7)	10.4 (1.6)	1.5 (1.3–1.9)	22.4 (3.6)	3.8 (2.9–5.1)	22.1 (8.0)	3.8 (2.3–6.1)	7.0 (0.6)
Any mental disorder	10.3 (0.9)	0.5 (0.5–0.6)	30.2 (2.6)	2.0 (1.8–2.3)	58.6 (4.3)	6.6 (5.2–8.5)	74.7 (6.2)	13.8 (8.6–22.0)	17.6 (1.0)
Women
Affective disorder	2.0 (0.3)	0.3 (0.2–0.3)	9.5 (1.1)	1.4 (1.1–1.7)	26.3 (3.0)	4.7 (3.8–5.7)	50.6 (5.7)	13.4 (9.8–18.3)	7.1 (0.5)
Anxiety disorder	9.1 (0.7)	0.5 (0.4–0.5)	26.7 (1.9)	1.7 (1.5–1.9)	45.2 (3.6)	3.8 (3.1–4.6)	72.7 (5.1)	12.2 (8.5–17.6)	17.9 (0.7)
Substance use disorder	2.1 (0.4)	0.7 (0.5–0.8)	4.2 (0.7)	1.3 (1.0–1.7)	7.4 (1.4)	2.4 (1.7–3.4)	9.8 (3.1)	3.2 (1.9–5.5)	3.2 (0.4)
Any mental disorder	11.7 (0.7)	0.5 (0.4–0.5)	33.5 (2.0)	1.8 (1.6–2.0)	56.2 (3.9)	4.5 (3.7–5.4)	82.7 (4.9)	16.6 (10.8–25.6)	22.3 (0.7)

SSLR, stratum-specific likelihood ratio.

Stratum-specific likelihood ratios are also presented in Table 3. The magnitudes of the SSLRs indicate that the odds of having each mental disorder class as well as any mental disorder increased across K10 strata. Most noticeably the odds of having all disorder groups except any substance use disorder and any mental disorder were greater than 10 for people in the very high K10 strata. This can be interpreted as meaning that there is sufficient evidence of the presence of any affective, any anxiety or any mental disorder in those with very high K10 scores. At the other end of the extreme none of the SSLRs were less than 0.1 indicating that the K10 does not provide sufficient evidence to rule out mental disorder. Comparing across the mental disorder classes within each K10 stratum the SSLRs are generally highest for any affective disorder, slightly lower for any anxiety disorder and lowest for any substance use disorder.

It should be noted that equivalent versions of all tables presenting data on the K6 are available as supplementary material. (available online at http://informahealthcare.com/doi/abs/10.3109/00048674.2010.543653).

Discussion

The findings of the current study represent the most up-to-date Australian normative data on the K10, a measure of psychological distress commonly used in large general population surveys. These data can now be used by clinicians and researchers alike to provide a general population benchmark against which to compare the levels of psychological distress in their individuals or samples of interest.

The current study confirms the often observed finding that psychological distress is higher in women than men. It has been suggested that cultural norms relating to the expression of distress may in part be driving observed differences in rates of psychological distress between women and men. However, a recent methodological study on the construct validity of the K6 (a shortened form of the K10 made up of six of the original ten items) concluded that the K6 is both a sex- and age-neutral scale [54]. This means that any observed differences in K6 scores between women and men or between people at different ages reflect true differences in psychological distress and are not the result of bias brought about by differential interpretation or understanding of the K6 questions.

The finding that K10 scores were higher in those with mental disorders, compared to those without, is entirely consistent with expectations and provides further evidence of the validity of the K10 as a screener for mental disorders. Moreover, scores on the K10 were more affected by the presence of a mental disorder than the presence of a physical condition. The SSLRs provide evidence of the increased or decreased likelihood of having a mental disorder given a score on the K10, and substantiate the strong link between K10 scores at the upper end of the range (i.e. in the very high strata) and the presence of mental disorders. This finding is again not unexpected, given that the K10 was specifically designed to provide maximum precision and discrimination at the 90th percentile of the general population [1].

The SSLRs presented in Table 3 are particularly useful as they provide information necessary to determine how the K10 score of a given individual is related to the probability of having a mental disorder. In order to see how this works it is necessary to define some terminology. In the absence of a K10 score a given individual's probability of having a mental disorder can be reasonably assumed to be the same as the population prevalence of a mental disorder. This is often referred to as the pre-test probability. With knowledge of an individual's K10 score the probability of having a mental disorder often changes (it may increase or it may decrease) and this new probability can be referred to as the post-test probability. This post-test probability is what researchers, clinicians and consumers themselves are most interested in as it can help in deciding whether to rule in or rule out a diagnosis.

Using the mathematical relationship of Bayes’ theorem [55] it can be shown:

Post-test odds of mental disorder = Pre-test odds of mental disorder × SSLR

where

Odds of mental disorder = Probability of mental disorder/(1 − Probability of mental disorder)

Based on this relationship, SSLRs generated in a benchmark sample such as a general population sample can be used to easily compute predicted probabilities of mental disorder from K10 scores obtained from other samples with prevalences different from those of the general population. This is perhaps best understood by example. Consider a research study in which the K10 is administered to a sample of people which has a known or assumed prevalence of any affective disorder of 40%. The question becomes how this pre-test probability of 40% is changed with knowledge that an individual's K10 score is in the very high range. Using the SSLRs reported in Table 3 and Bayes’ theorem we can calculate an individual's post-test probability of having an affective disorder in the following way:

The calculations shown above illustrate how a prior probability of mental disorder of 0.4 is updated to a probability of 0.92 with knowledge that an individual has scored in the very high range on the K10. While it is possible to carry out the above calculations to determine a post-test probability for any exact pre-test probability, Figure 1 provides the pattern of post-test probabilities of mental disorder by K10 strata across nine pre-test probabilities ranging from 0.1 to 0.9. Arrows are shown in the top left graph (any affective disorder) to illustrate how to determine the post-test probability directly from the graph. The pattern of results presented in Figure 1 confirms the superior performance of the K10 as a screener for mental disorder, particularly for any affective disorder. However, perhaps its main benefit is to provide easily calculable clinical information regarding the extent of psychological distress in a given individual or research sample.

Figure 1.

Post-test probabilities of any affective disorder (A), any anxiety disorder (B), any substance use disorder (C) and any mental disorder (D) by K10 strata across nine pre-test probabilities ranging from 0.1 to 0.9.

In conclusion, we have presented the latest normative data on the K10 in an Australian community sample. The K10 is a scale widely used not only as an indicator of mental health in the general population but also as a screening tool to detect the presence of mental disorder in individuals. We hope these data will be useful across a broad range of applications.

Footnotes

Acknowledgements

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Supplementary material available online

Appendix questionnaire

References

1. Kessler

Andrews

Colpe

. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med 2002; 32:959–976.

2. Kessler

Mroczek

. Final versions of our Non-Specific Psychological Distress Scale [memo dated 10/3/94]. Ann Arbor(MI): Survey Research Center of the Institute for Social Research, University of Michigan, 1994.

3. Kessler

Mroczek

. An update of the development of mental health screening scales for the us national health interview study [memo dated 12/22/92]. Ann Arbor(MI): Survey Research Center of the Institute for Social Research, University of Michigan, 1992.

4. Hides

Lubman

Devlin

. Reliability and validity of the Kessler 10 and Patient Health Questionnaire among injecting drug users. Aust N Z J Psychiatry 2007; 41 (2):166–168.

5. Lacey

Salzberg

Roberts

Trauer

D'Souza

. Psychiatric comorbidity and impact on health service utilization in a community sample of patients with epilepsy. Epilepsia 2009; 50:1991–1994.

6. Spies

Kader

Kidd

. Validity of the K-10 in detecting DSM-IV-defined depression and anxiety disorders among HIV-infected individuals. AIDS Care 2009; 21:1163–1168.

7. Nagel

Robinson

Condon

Trauer

. Approach to treatment of mental illness and substance dependence in remote Indigenous communities: results of a mixed methods study. Aust J Rural Health 2009; 17:174–182.

8. Berry

Rodgers

. Trust and distress in three generations of rural Australians. Australasian Psychiatry 2003; 11:S131–S137.

9. Eckert

Taylor

Wilkinson

Tucker

. How does mental health status relate to accessibility and remoteness? Med J Aust 2004; 181:540–543.

10.

10. Fragar

Stain

Perkins

. Distress among rural residents: does employment and occupation make a difference? Aust J Rural Health 18:25–31.

11.

11. Kelly

Stain

Coleman

. Mental health and well-being within rural communities: the Australian Rural Mental Health Study. Aust J Rural Health 2010; 18:16–24.

12.

12. Kilkkinen

Kao-philpot

O'Neil

. Prevalence of psychological distress, anxiety and depression in rural communities in Australia. Aust J Rural Health 2007; 15:114–119.

13.

13. Lee

Browne

. Subjective well-being, sociodemographic factors, mental and physical health of rural residents. Aust J Rural Health 2008; 16:290–296.

14.

14. Stain

Kelly

Lewin

Higginbotham

Beard

Hourihan

. Social networks and mental health among a farming population. Soc Psychiatry Psychiatr Epidemiol 2008; 43:843–849.

15.

15. Donker

Comijs

Cuijpers

. The validity of the Dutch K10 and extended K10 screening scales for depressive and anxiety disorders. Psychiatry Res 2009; 176:45–50.

16.

16. Furukawa

Kawakami

Saitoh

. The performance of the Japanese version of the K6 and K10 in the World Mental Health Survey Japan. Int J Methods Psychiatr Res 2008; 17: 152–158.

17.

17. Baggaley

Ganaba

Filippi

. Detecting depression after pregnancy: the validity of the K10 and K6 in Burkina Faso. Trop Med Int Health 2007; 12:1225–1229.

18.

18. Cairney

Veldhuizen

Wade

Kurdyak

Streiner

. Evaluation of two measures of psychological distress as screeners for depression in the general population. Can J Psychiatry 2007; 52:111–120.

19.

19. Donker

van Straten

Marks

Cuijpers

. Brief self-rated screening for depression on the Internet. J Affect Disord 2009; 122:253–259.

20.

20. Mackinnon

Jorm

Hickie

. A National Depression Index for Australia. Med J Aust 2004; 181S7:S52–S56.

21.

21. Patel

Araya

Chowdhary

. Detecting common mental disorders in primary care in India: a comparison of five screening questionnaires. Psychol Med 2008; 38:221–228.

22.

22. Silove

Bateman

Brooks

. Estimating clinically relevant mental disorders in a rural and an urban setting in postconflict Timor Leste. Arch Gen Psychiatry 2008; 65:1205–1212.

23.

23. Tang

GWG

Dennis

Comino

Zwar

. Anxiety and depression in Chinese patients attending an Australian GP clinic. Aust Fam Physician 2009; 38:552–555.

24.

24. Tesfaye

Hanlon

Wondimagegn

Alem

. Detecting postnatal common mental disorders in Addis Ababa, Ethiopia: validation of the Edinburgh Postnatal Depression Scale and Kessler Scales. J Affect Disord 2010; 122:102–108.

25.

25. Baillie

. Predictive gender and education bias in Kessler's psychological distress Scale (K10). Soc Psychiatry Psychiatr Epidemiol 2005; 40:743–748.

26.

26. Spies

Stein

Roos

. Validity of the Kessler 10 (K-10) in detecting DSM-IV defined mood and anxiety disorders among pregnant women. Arch Women's Ment Health 2009; 12:69–74.

27.

27. Arnaud

Malet

Teissedre

. Validity study of Kessler's Psychological Distress Scales conducted among patients admitted to French emergency departments for alcohol consumption-related disorders. Alcohol Clin Exp Res 2010; 34:1235–1245.

28.

28. Wittchen

H-U

. Screening for serious mental illness: methodological studies of the K6 screening scale. Int J Methods Psychiatr Res 2010; 19 (S1):1–3.

29.

29. Centre for Epidemiology and Research. Summary report on adult health from the New South Wales Population Health Survey. Sydney: NSW Department of Health, 2006–2008.

30.

30. Centre for Population Studies in Epidemiology. South Australian Health and Wellbeing Survey. South Australia: Department of Human Services, 2000.

31.

31. Health and Wellbeing Survey 2000: psychological distress in the Western Australian population. Perth: Health Department of Western Australia, 2001. [Cited 10 August 2010.] Available from URL http://www.health.wa.gov.au/publications/documents/distress.pdf.

32.

32. State Government of Victoria. Victorian Population Health Survey. Department of Human Services; 2001–2007. [Cited 20 August 2010.] Available from URL http://www.health.vic.gov.au/healthstatus/vphs_current.htm.

33.

33. Australian Bureau of Statistics. National Health Survey: summary of results 2001. Canberra: Australian Bureau of Statistics, 2002.

34.

34. Australian Bureau of Statistics. National Health Survey: summary of results 2004–2005. Canberra: Australian Bureau of Statistics, 2006.

35.

35. Australian Bureau of Statistics. National Health Survey: summary of results 2007–2008. Canberra: Australian Bureau of Statistics, 2009.

36.

36. Mental health national outcomes and casemix collection: technical specification of state and territory reporting requirements, version 1.6. Canberra: Department of Health and Ageing, 2009.

37.

37. Pirkis

Bassilios

Fletcher

. Clinical improvement after treatment provided through the Better Outcomes in Mental Health Care program: do some patients fare better than others? Aust N Z J Psychiatry, in press.

38.

38. Andrews

Slade

. Interpreting scores on the Kessler Psychological Distress Scale (K10). Aust N Z J Public Health 2001; 25:494–497.

39.

39. Kessler

Green

Gruber

. Screening for serious mental illness in the general population with the K6 screening scale: results from the WHO World Mental Health (WMH) survey initiative. Int J Methods Psychiatr Res 2010; 19S1:4–22.

40.

40. Furukawa

Kessler

Slade

Andrews

. The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health and Well-Being. Psychol Med 2003; 33:357–362.

41.

41. Australian Bureau of Statistics. National Survey of Mental Health and Wellbeing: user's guide, 2007. Canberra: Australian Bureau of Statistics, 2009.

42.

42. Slade

Johnston

Oakley Browne

Andrews

Whiteford

. 2007 National Survey of Mental Health and Wellbeing: methods and key findings. Aust N Z J Psychiatry 2009; 43:594–605.

43.

43. Kessler

Üstün

. The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). Int J Methods Psychiatr Res 2004; 13:93–121.

44.

44. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, fourth. Washington DC: American Psychiatric Association, 1994.

45.

45. World Health Organization. International statistical classification of diseases and related health problems, tenth. Geneva: World Health Organization, 1992.

46.

46. Haro

Arbabzadeh-Bouchez

Brugha

. Concordance of the Composite International Diagnostic Interview version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health Surveys. Kessler

Üstün

. The WHO World Mental Health Surveys: global perspectives on the epidemiology of mental disorders, Chapter 6, Part I: Methods. New York: Cambridge University Press, 2008: 114–127.

47.

47. Kessler

Abelson

Demler

. Clinical calibration of DSM-IV diagnoses in the World Mental Health (WMH) version of the World Health Organization (WHO) Composite International Diagnostic Interview (WMH-CIDI). Int J Methods Psychiatr Res 2004; 13:122–139.

48.

48. SAS Proprietary Software 9.2. Cary, NC: SAS Institute, 2008.

49.

49. Stata Version 10. College Station, TX: StataCorp, 2007.

50.

50. Rust

Rao

. Variance estimation for complex surveys using replication techniques. Stat Methods Med Res 1996; 5:283–310.

51.

51. Australian Bureau of Statistics. Use of the Kessler Psychological Distress Scale in ABS health surveys. Information paper. Canberra: Australian Bureau of Statistics, 2001.

52.

52. Peirce

Cornell

. Integrating stratum-specific likelihood ratios with the analysis of ROC curves. Med Decis Making 1993; 13:141–151.

53.

53. Jaescke

Guyatt

Lijmar

. Diagnostic tests. Guyatt

Rennie

. User's guides to the medical literature. Chicago: AMA Press, 2002: 121–140.

54.

54. Drapeau

Beaulieu-Prevost

Marchand

Boyer

Preville

Kairouz

. A life- course and time perspective on the construct validity of psychological distress in women and men. Measurement invariance of the K6 across gender. BMC Med Res Methodol 2010; 10:68.

55.

55. Bayes

. An essay towards solving a problem in the doctrine of chances. Philos Trans R Soc Lond 1793; 53:370–418.

Kessler Psychological Distress Scale: Normative Data from the 2007 Australian National Survey of Mental Health and Wellbeing

Abstract

Keywords

Method

Sample and measures

Statistical analysis

Results

Discussion

Footnotes

Acknowledgements

Supplementary material available online

References