Monitoring Patient Treatment Outcomes in Inpatient Psychiatric Care with the PHQ-9

December 2, 2020
Rachel B. Nowlin, MS

,
Sarah K. Brown, DrPH

The PHQ-9 may be an effective asset in a clinician’s toolkit, along with their clinical judgment and therapeutic alliance, to ensure treatment planning and outcome tracking is personal to each patient.

As more behavioral health organizations embrace the use of patient outcomes in treatment planning and quality measurement, the goal becomes identifying appropriate measures. Ideally, measurements would be brief, easy-to-use, and sensitive to patients’ concerns. For inpatient psychiatric settings, one such assessment may be the Patient Health Questionnaire (PHQ-9), as it is only 9 items, open-domain, patient self-report, and addresses mental health.

Given its purpose as a screener for depression in primary-care settings, the PHQ-9 has attracted interest as a measure of treatment outcomes in psychiatric settings.1 Research suggests the PHQ-9 can detect change in depression symptoms in a medical population and in outpatient psychiatric settings,2,3 but it has not been validated as a measure of symptom improvement in acute inpatient psychiatric care.

One pitfall preventing validation of the PHQ-9 in inpatient settings may be the timeframe of symptom evaluation on the measure. On the typical PHQ-9, patients are asked to respond how frequently an item has been a concern in the last 2 weeks, ranging from “Not at all” to “Nearly every day.” However, inpatient stays in acute settings are rarely even 2 weeks in length, so this timeframe could undercut capturing change since admission. We propose an alternative evaluation period of 1 week, to fall within the length of a modern inpatient stay. Other studies have altered the timeframe of the PHQ-9 to address research questions,4,5 but routine use of the PHQ-9 with an altered timeframe has not been examined previously.

This analysis provides an example for use of a shortened-timeframe (1 week) PHQ-9 as a measure of outcomes for adult and geriatric patients in acute psychiatric inpatient treatment. We demonstrate that the PHQ-9 is far from limited in its applicability and can detect symptom improvement for patients—including those with diagnoses other than depression—within the span of an inpatient stay.

Methods and Participants 

Admission and discharge PHQ-9 records were available for 27,991 adult and geriatric encounters from facilities participating in routine outcomes monitoring and benchmarking for acute psychiatric inpatient programs. The data represented 41 geographically diverse facilities across 2018 and 2019. For this sample, 43.5% of patients had a primary diagnosis of major depressive disorder (MDD). Average length of stay for the overall sample was 8.8 days, just over the modified PHQ-9’s 1 week evaluation period.

Key Findings

1 . The PHQ-9 can categorize depression in an acute inpatient psychiatric setting

  • At admission, average depression severity falls between the moderate and moderately severe category cutoffs for the total sample (14.6 of a possible 27 points, with higher scores indicating greater symptom severity). Patients with a primary diagnosis of MDD and patients without primary MDD scored in these category ranges (15.7 and 13.7 respectively).
  • By discharge, the overall sample’s average PHQ-9 severity was between mild and minimal ranges (score of 4.9). Both patients with a primary diagnosis of MDD and those without scored in these ranges (5.1 and 4.8 respectively).
  • As anticipated, patients with a primary diagnosis of MDD reported greater severity at both admission and discharge compared to non-MDD patients. Although those differences were significant, the effect is too small for meaningful conclusions.

2. The PHQ-9 can detect change in depressive symptomology

  • At discharge, the average decrease in severity indicates a significant and meaningful change (9.7 points) from admission to discharge (p<.001, Cohen’s d=1.155). Patients with a primary diagnosis of MDD reported a higher average change score compared to those without (10.6 and 9.0 respectively).
  • Change scores were evaluated using a reliable change index, revealing the majority (76.8%) of patients’ symptom improvement were statistically reliable. A higher proportion of patients with a primary diagnosis of MDD demonstrated reliable improvement than non-MDD patients (81.5% versus 73.1% respectively; p<.001).

3. The PHQ-9 captured depressive symptomology for most, but not all patients:

  • Roughly 6% of all patients in the sample reported no depression symptoms at admission (ie, a PHQ-9 score of 0), and the proportion was higher in the non-MDD group (8.4%) than the MDD group (3.7%).
  • 4.4% of all patients reported a 0 at both admission and discharge. This group of patients showed 2 notable differences from other patients: a higher rate of primary diagnosis of anxiety disorders (including adjustment disorders and PTSD) than other patients and a shorter length of stay.

Conclusions

Our results suggest a modified 1 week timeframe PHQ-9 is effective at registering depression severity at admission and measuring change during a stay for most adult and geriatric psychiatric inpatients. The PHQ-9 may be tailored for people with depression, but depression symptoms overlap with many disorders, and there is often significant comorbidity in an acute setting.

Based on these findings, we believe the PHQ-9 is a broadly useful tool for treatment outcome measurement in psychiatric inpatient units. As is true with any measure, the PHQ-9 will not serve as the best measure of outcomes for all patients. There will be patients who report no difficulty with the PHQ-9 items, patients who report no improvement, or, unfortunately, patients who report greater difficulty at discharge. A patient may see an increase in depression symptoms while addressing addiction or trauma, for example.

As such, the PHQ-9 may be an effective asset in a clinician’s toolkit, along with their clinical judgment and therapeutic alliance to ensure treatment planning and outcome tracking is personal to each patient. 

Ms Nowlin and Dr Brown are data analysts for Mental Health Outcomes, LLC., a subsidiary of Universal Health Services, Inc, Lewisville, TX.

References

1. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606-613.

2. Löwe B, Kroenke K, Herzog W, Gräfe K. Measuring depression outcome with a brief self-report instrument: Sensitivity to change of the Patient Health Questionnaire (PHQ-9). J Affect Disord. 2004;81:61-66.

3. Beard C, Hsu KJ, Rifkin LS, Busch AB, Björgvinsson T. Validation of the PHQ-9 in a psychiatric sample. J Affect Disord. 2015;193(2016):267-273.

4. Cannon DS, Tiffany ST, Coon H, Scholand MB, McMahon WM, Leppert MF. The PHQ-9 as a brief assessment of lifetime major depression. Psychol Assess. 2007;19(2):247-251.

5. Feng Y, Huang W, Tian T-F, et al. The psychometric properties of the Quick Inventory of Depressive Symptomatology-Self-Report (QIDS-SR) and the Patient Health Questionnaire-9 (PHQ-9) in depressed inpatients in China. Psychiatry Res. 2016;243:92-96.