Chart 5 Sample Report for Psychosis Symptom Severity
The measures discussed in the previous chapters have all been self-report measures. This chapter will discuss six clinician-rated instruments developed by the DSM-5 authors that can be used to evaluate behaviors found across different disorders. These measures, as with the others found on the DSM-5 website, are in the public domain and can be used for clinical purposes without violating any copyright laws. They can be used to monitor changes in frequency and duration of specific symptoms, or changes in behavior. They may also be useful for monitoring changes when medications are introduced or changed.
As a psychologist, part of my basic training and ongoing professional development was in psychological assessment. It is useful to remember that training in administering, scoring, and interpreting psychological tests and measures is NOT typically part of medical education, and only a small part of training in most master’s level psychology programs. At first glance, it may appear that it is easy to administer and score the clinician-rated measures contained in the DSM-5, but I would challenge that assumption and encourage all clinicians to follow the applicable ethical guidelines regarding limitations in administering or interpreting tests and measures that they may be subject to.
Going to the DSM5.org website and clicking the link for Online Assessment Measures under “What’s New” and scrolling approximately two-thirds of the way down will bring you to six clinician-rated instruments. These are diagnostically useful, brief rating scales specific to several disorders: oppositional defiant disorder, conduct disorder, autism spectrum and social communication disorders, psychosis symptom severity (also available in print book), somatic symptom disorder, and non-suicidal self-injury. Each instrument contains instructions on scoring and frequency of use. They do not include interpretation of results.
With minor variations within a specific instrument, all these are scored on a 5-point Likert scale, with the clinician exercising clinical judgment in assigning levels of severity. Unlike the client self-report measures, there is no weighting of individual questions or summing of domains to determine overall severity. Each item should be viewed as separate and distinct, and used for designing specific interventions when treatment planning.
One of the challenges in using clinician-rated scales is that without some training to insure inter-rater reliability, one person’s “extreme” is another’s “mild”. To address this variability, the DSM-5 provides thumbnail descriptors for each of the Likert-points. Even with that, however, these measures are subject to great variability in assignment of rankings. Psychologists recognize this and will typically use several measures to identify and explore issues. There is no such recommendation in the DSM-5, which is a threat to the overall utility of this information. As a provider, it would be prudent to make sure you are evaluating not just the symptoms, but the overall functioning of your client. Thus you may want to refer the client to a psychologist for additional testing or, at a minimum, include several measures (self-report and clinician-rated) to improve the validity of the information.
The dilemma here lies more with inter-rater reliability than test-retest issues. If the measure is used as an inpatient monitor, it is essential that all clinicians who will be using it are trained on how to assess and evaluate patients using this measure. There may be more problems with outpatient evaluation as there is no specific training, and the evaluations will differ based on the clinician’s level of training and experience with the disorder. As noted above, while the measures provide clear descriptions of the behaviors, there still is room for variability in evaluating severity. I believe we develop a type of clinical tolerance based on our backgrounds and clinical exposure to certain behaviors. What may be extreme behavior to a newly licensed clinician who has had little exposure to an individual experiencing a full-blown psychotic episode, may be “mild” or “moderate” to a clinician who has been practicing for several years in an in-patient setting. We need to remember that these are quantitative estimations of qualitative experiences.
A typical arc of use would include identifying the presence of the symptoms (baseline), and then following the course of the episode to assess reduction in or absence of symptoms over time. Since these measures are clinician-rated, there is no test-retest problem. In an outpatient setting, the clinician can use the measures on a weekly basis or less frequently. You may wish to do an initial evaluation (baseline), then follow up in several weeks. Or you may want to track specific symptoms for several weeks to see if there are any changes. In an inpatient setting, these measures can be used at each shift, daily, or weekly, depending on need and benefit.
Charting could be as simple as using the measure, filling in the boxes and putting the measure in the chart. The DSM-5 forms do not contain a space for the name of the clinician who is making the rating. This is an important oversight and should be addressed in your integration of these measures. If you are using an electronic health record (EHR), you can include these measures and the ratings within the program itself. There may be costs involved in uploading these and scoring them depending on which EHR you are using, so you should consult with the developer.
The table below lists each of the measures. They are available at www.dsm5.org under “Online Assessment Measures”.
Table 2 Clinician-Rated Severity Measures for specific disorders and conditions
The reader will note that most of these are one- or two-item measures. They ask the clinician to use all available information and clinical judgment to accurately select a level that is representative of the client’s functioning for that particular item. Evaluation periods differ amongst the disorders ranging from the past 7 days to an entire year. Measures are completed at assessment. They are best used for treatment planning and prognosis, not diagnosis.
Each instrument looks at something a bit different and is noted in the table below.
Table 3 Specific Domains of Assessment in Clinician-Rated Measures
| Area assessed | Assesses the . . . |
| Autism | level of interference in functioning and support required as a result of difficulties in 1) social communication and 2) restricted interests and repetitive behaviors that are present for the individual receiving care |
| Somatic Symptoms | severity of the individual’s misattributions, excessive concerns, and/or preoccupations with the somatic symptom(s) |
| Oppositional Defiant | presence and severity of any oppositional defiant symptoms |
| Conduct Disorder | presence and severity of any conduct disorder symptoms |
| Non-suicidal Self-injury | presence and severity of any non-suicidal self-injury (NSSI) behaviors or problems |
One of the philosophical and theoretical differences between the DSM-IVTR and the DSM-5 is the idea of heterogeneity of disorders. In the DSM-5 world, psychosis appears in any number of different disorders (e.g., schizophrenia, bipolar disorder, substance use, delirium), and manifests itself differently in terms of severity and impairment in functional domains. The Clinician-Rated Dimensions of Psychosis Symptom Severity Scale is a useful tool to look at how psychosis manifests itself across diagnostic categories.
The instrument evaluates eight domains of functioning. These comprise positive symptoms (hallucinations, delusions, disorganized speech, and abnormal psychomotor behavior), negative behaviors, as well as impaired cognition, depression, and mania. Each of these domains is evaluated using a 5-point Likert scale, with specific examples of typical manifestations of the behaviors across the severity spectrum. Thus, you can evaluate the client’s functioning over the past seven days looking at more than just whether a specific symptom is absent or present, and you can monitor overall functioning noting which domain(s) are more impacted. This allows clinicians to track the incredible variability of psychotic behavior in patients.
The scale goes from “0” (not present), “1” (equivocal), “2” (present but mild), “3” (present and moderate), and “4” (present and severe). These scales include descriptors for behaviors within each of the specific domains. For example, the descriptors for hallucinations (a positive symptom) include the following:
Table 4 Severity descriptions for Hallucinations in Clinician-Rated Psychosis Symptoms
| 0 Not present | 1 Equivocal (severity or duration not sufficient to be considered psychosis) | 2 Present, but mild, (little pressure to act upon voices, not very bothered by voices) | 3 Present and moderate (some pressure to respond to voices, or is somewhat bothered by voices) | 4 Present and severe (severe pressure to respond to voices, or is very bothered by voices) |
Reprinted with permission from the Diagnostic and Statistical Manual of Mental Health Disorders, Fifth Edition, (Copyright © 2013). American Psychiatric Association. All Rights Reserved.
Descriptors for negative symptoms are:
Table 5 Severity descriptions for Negative Symptoms in Clinician-Rated Psychosis Symptoms
| 0 Not present | 1 Equivocal decrease in facial expressivity, prosody, gestures, or self-initiated behavior | 2 Present, but mild, decrease in facial expressivity, prosody, gestures, or self-initiated behavior | 3 Present and moderate decrease in facial expressivity, prosody, gestures, or self-initiated behavior | 4 Present and severe decrease in facial expressivity, prosody, gestures, or self-initiated behavior |
Reprinted with permission from the Diagnostic and Statistical Manual of Mental Health Disorders, Fifth Edition, (Copyright © 2013). American Psychiatric Association. All Rights Reserved. Impaired cognition is evaluated according to standard deviations1
Table 6 Severity descriptions for Impaired Cognition in Clinician-Rated Psychosis Symptoms
| 0 Not present | 1 Equivocal (cognitive function not clearly outside the range expected for age or SES; i.e., within 0.5 SD of mean) | 2 Present, but mild, (some reduction in cognitive function; below expected for age and SES, 0.5–1 SD from mean) | 3 Present and moderate (clear reduction in cognitive function; below expected for age and SES, 1–2 SD from mean) | 4 Present and severe (severe reduction in cognitive function; below expected for age and SES, > 2 SD from mean) |
Reprinted with permission from the Diagnostic and Statistical Manual of Mental Health Disorders, Fifth Edition, (Copyright © 2013). American Psychiatric Association. All Rights Reserved.
The chart below shows data collected on a patient with schizoaffective disorder who was a resident of a skilled nursing facility. I tracked his symptoms over a six-week period during which time his medications were being adjusted. The adjustments were successful, and the scale captures the decrease in depression, negative behaviors, agitation, and hallucinations. This information was shared with his psychiatrist, who was able to follow the patient more closely. It also provided nursing home staff with information on specific issues related to the patient’s interpersonal functioning that helped them in supporting the patient in engaging in healthier behaviors. No other clinician was rating this individual.
Data can be gathered over time and used to monitor changes in behavior and effectiveness of interventions.
Chart 5 Sample Report for Psychosis Symptom Severity
Because the nursing staff saw improvement, I was asked to do an in-service training on psychosis in residents with mild and major neurocognitive disorders, personality disorders, and depression. This demonstrated to the facility staff that these behaviors occurred in a variety of organic disorders, not just “mental illness.” Nurses were encouraged to use the scale to rate their patients and make referrals to behavioral health staff. Additional training was provided to the nursing staff in order to improve inter-rater reliability. The measure was uploaded to the EHR and the nursing staff would enter their rating along with other vital signs.
Historically, these types of data would have been summarized and shared with other professionals, but most likely not with the clients. My personal experience is that sometimes this information is useful to clients, and sometimes it is not. This invites clinicians to use our clinical judgment in sharing these results. You should also note, however, that the recent changes to the HIPAA law have expanded access to this type of information for clients. Depending on laws within your state, there may be additional rights extended to clients for obtaining this information.
Rather than state unequivocally that you should or should not provide this information, I would encourage you to consult with your local provider group, risk management (malpractice) providers, supervisors, and other similar resources to identify what is required in your specific locale. My personal practice is to share this information on a case-by-case basis in the interest of providing both information and insight to my clients. This may require my spending time explaining not just the results, but how they are determined and interpreting what these things mean. For the most part, this is usually not reimbursable.
This information may also be useful for other providers, especially primary care providers (e.g., physicians, nurse practitioners) who typically do not have a lot of time for narrative or explanation. Finding ways to succinctly state what the results show is important, and may be useful in the overall continuum of care for the client. It is important to reiterate that just sharing raw data with providers is not likely to be useful.
Clinician-rated scales are useful for monitoring specific symptoms over time. With proper training and collective understanding and agreement of what each of the descriptors means, clinicians can easily gather information and then use it to track the effectiveness of their interventions, adapt treatment plans, and chart overall changes in the client. These measures are vulnerable to several important areas of reliability. Inter-rater reliability may be low without sufficient training on the part of the individual who is making the rating and without sufficient training among individuals using the measure within an in-patient or group setting. Test-retest reliability is dependent on how well the staff have been trained to administer the measures.
Use of these monitors can extend from just keeping the ratings in the chart, to producing comparisons of changes over time. Incorporating this information and sharing it between providers is valuable. It may also be valuable to share this information with clients. This is an area for clinical judgment and should be determined on a case-by-case basis.
______________________
1 Here the standard deviation refers to the normal distribution of individuals. Thus, the mean is reflecting distribution of individuals with cognitive problems within the population. One standard deviation above or below the mean suggests that the client is already experiencing challenges, and should not be interpreted as “normal”. As the standard deviations increase, the level of severity increases. Two standard deviations above the mean (if you recall from your statistics classes) represents the outer tail of the bell-shaped curve. This represents approximately 2.1% of the population and is, therefore, extreme or rare.