Screening for and subsequent participation in a trial for depression and anxiety in people with type 2 diabetes treated in primary care: Who do we reach?

AIMS

This study investigated (factors related to) (a) the response to a screening procedure for depression and anxiety in people with type 2 diabetes in primary care, and (b) participation in a subsequent randomised controlled trial targeting depressive or anxiety symptoms.

METHODS

People with type 2 diabetes (n=1837) received a screening questionnaire assessing depressive symptoms (PHQ-9) and anxiety symptoms (GAD-7). Eligible persons who scored above the cut-off score (PHQ-9≥7 or GAD-7≥8) were offered to participate in the trial.

RESULTS

In total, 798 people (43%) returned the screening questionnaire. Non-responders were more often female (53% vs 44%, p<0.001), had higher LDL cholesterol levels (Cohen's d=0.17, p=0.001) and a higher albumin/creatinine ratio (Cohen's d=0.08, p=0.01). In total, 130 people (18%) reported elevated depressive or anxiety symptoms. Twenty-seven persons agreed to participate in the trial. Factors related to participation were a high education level, a higher level of diabetes distress and a history of psychological problems.

CONCLUSIONS

Using screening as recruitment resulted in a small number of participants in a treatment trial for anxiety and depression. Research is needed to investigate whether screening is also followed by a low uptake of treatment in primary care outside a RCT setting.

Effectiveness of a stepped care intervention for anxiety and depression in people with diabetes, asthma or COPD in primary care: A randomized controlled trial

BACKGROUND

Depression and anxiety are common in people with a chronic somatic disease. Although guidelines recommend stepped care, the effectiveness of this approach has not been evaluated in people with diabetes, asthma, or COPD in primary care.

METHODS

3559 People were sent screening questionnaires (41% response). Of 286 persons with anxiety and/or depression (Generalized Anxiety Disorder questionnaire, GAD-7, cut-off ≥ 8 and/or Patient Health Questionnaire, PHQ-9, cut-off ≥ 7), 46 were randomized into the intervention (stepped care and monitoring of symptoms; n = 23) or control (usual care) group (n = 23). Main outcomes were symptoms of anxiety and depression after the 12-months intervention and six months post intervention. Analysis of covariance was first adjusted for condition and baseline GAD-7/PHQ-9 scores and additionally for age, sex and education.

RESULTS

The intervention group had a significantly lower level of anxiety symptoms at the end of the program (GAD-7 6 ± 6 vs. 9 ± 6; Cohen's d = 0.61). This effect was still present six months post intervention. The effect on depression was statistically significant in the first model (PHQ-9 6 ± 4 vs. 9 ± 6; p = 0.035), but not in the fully adjusted model (p = 0.099), despite a large effect size (d = 0.63). At six months post intervention there was no statistically significant difference in symptoms of depression between the two groups although the difference in symptoms was still clinically significant (Cohen's d = 0.61).

LIMITATIONS

Many people were screened, but relatively few participated in the randomized controlled trial.

CONCLUSIONS

Stepped care with monitoring resulted in a lower level of symptoms of anxiety and depression in people with a chronic condition.

Diabetes-specific emotional distress in people with Type 2 diabetes: a comparison between primary and secondary care

AIMS

To compare levels of diabetes distress in people with Type 2 diabetes treated in primary and secondary care and to examine demographic and clinical correlates that may explain potential differences in levels of distress between care settings.

METHODS

People with Type 2 diabetes from 24 primary care practices (n = 774) and three secondary care clinics (n = 526) completed the Problem Areas In Diabetes questionnaire. Data on HbA1c levels and diabetes complications were derived from medical charts. Hierarchical ordinal regression analysis was used to investigate which correlates could explain the potential differences in level of diabetes distress between care settings.

RESULTS

Diabetes distress levels and the prevalence of elevated diabetes distress were considerably lower in the participants treated in primary care (mean (SD) total diabetes distress score 8 (11); 4% of participants with a Problem Areas In Diabetes score ≥ 40) than in secondary care (mean (SD) total diabetes distress score 23 (21); 19% of participants with a Problem Areas In Diabetes score ≥ 40, P < 0.001). In addition to care setting, the following variables were also independently related to diabetes distress: younger age, ethnic minority status, using insulin, having a higher HbA1c level, having a higher BMI and the presence of neuropathy. Other diabetes complications were not independently associated with diabetes distress.

CONCLUSIONS

In primary care, lower levels of diabetes distress were reported than in secondary care. The difference in diabetes distress between care settings can be largely, but not fully, explained by specific demographic and clinical characteristics. These results need to be interpreted with caution as they are based on two separate studies, but do call into question the need to screen for diabetes distress in people with Type 2 diabetes in primary care.