Is easier and quicker… better?
In recent years, targeted or single-domain assessments (i.e., only ruling in/out one diagnosis as opposed to a comprehensive evaluation) have increased significantly, especially with Telehealth and insurance-based evaluations, as well as being offered by clinicians who may only have niche or limited testing background (e.g., a school psychologist testing for learning disabilities or a psychologist with minimal neuropsychological training assessing for ADHD via observations and questionnaires). Targeted evaluations may result in potential bias of clinicians, psychological tests, and diagnostic criteria.
One of the big challenges and concerns of a targeted evaluation is the significant overlap in presentations and experiences of “different” disorders. Presently, the DSM is a categorical-based system. That is, essentially, you either have it or you don’t. Unfortunately, there are many downsides to this method of classification. Part of the confusion lies with the fact that there are so many different combinations (hundreds or thousands for some disorders with multiple potential criteria) that result in the same diagnosis with a criteria-based approach, not to mention high rates of comorbidity across disorders, leading to significant heterogeneity across individuals with the same diagnosis. As such, there is often significant variability in presentation within the same diagnosis. Think about people you know who may carry an ADHD, autism, or depression diagnosis and how differently they may present in terms of their own strengths and struggles. Further, from a neuropsychological perspective, the same diagnosis can result in significant differences in skill sets. For example, for children for whom the best diagnosis would be ADHD, that could mean many things: extremely slow processing speed OR extremely high processing speed, better selective attention than sustained OR vice versa, significant restlessness OR daydreaming). You probably get the point. Moreover, if you think of a disorder like autism, which is largely diagnosed via behavioral observation of symptoms and the presence of developmental delays, there can be many different “causes” including underlying genetic conditions and/or other causes resulting in “different wiring” so to speak.
How do we know this to be true, you might ask? Well, one study found that a low number of criteria across the DSM-5 mapped onto specific diagnoses (e.g., inattention, insomnia, avoidance, withdrawal), and almost 63% of criteria mapped to more than one disorder. The heterogeneity of symptom profiles was almost as high within a disorder group (i.e., people who theoretically have the same disorder) as between 2 disorder groups (i.e., people who are alleged to have different disorders) and no individual disorder group was separable from randomly selected groups of individuals with at least one disorder. Another study found that DSM-5 disorder criteria failed to separate individuals by symptom profiles any better than random assignment. As stated in the article, “rather than representing a method of separating groups based on symptoms, disorder labels primarily serve to emphasize a particular subset of symptoms without consideration of the entire symptom profile.”
Most targeted or single-diagnosis evaluations are focused on autism, ADHD, or a specific learning disability (i.e., dyslexia, dyscalculia, dysgraphia). The concern with evaluating for a single learning disability is that you may not only miss other closely associated learning disabilities or barriers to learning (e.g., developmental coordination disorder, language disorder) but you may also miss the true underlying cause of a learning difficulty or weakness, which may be better accounted for by anxiety, undiagnosed language delay, ADHD, and/or trauma, amongst others. In fact, one study found that while there is no gene or set of genes underlying risk for all disorders, subsets of disorders (e.g., depression, anxiety, OCD, anorexia) share a common genetic architecture. The high comorbidity across some disorders in part reflects overlapping pathways of genetic risk. For example, 70% of the genetic signal associated with schizophrenia is also associated with bipolar disorder (and as it stands you cannot be diagnosed with both). Moreover, the study also highlighted that anorexia and OCD had a strong shared genetic architecture, as did depression and anxiety. Additionally, there is an established connection between auditory processing disorder (APD), ADHD, and dyslexia. Those with co-occurring APD and dyslexia often struggle with sound discrimination and gap detection, both of which overlap with executive functioning deficits (common in both dyslexia and ADHD).
Let’s take autism spectrum disorder (ASD) for example. We know that there are an estimated 1100 genes implicated in ASD, many of which overlap with other neurodevelopmental disorders. Additionally, some genetic mutations result in genetic disorders, producing “global developmental delays” and “autism-like symptoms.” Which again begs the question, is this all one disorder? Does one label account for all these profiles and presentations? How many children are given an autism diagnosis in order to be eligible for services that are appropriate or do they truly fit the “autism” mold? A recent article compared the biomarkers and neuroanatomy of autism and ADHD and found that similar parts of the brain were affected in both disorders. Some have even postulated that ASD may be an “extreme form” of ADHD so to speak, as both often involve language, social-emotional, and sensory functioning deficits. In fact, and this is pretty significant, there were more shared biomarkers between Level 1 ASD and ADHD (which are supposedly two different disorders) than Level 1 ASD and Level 3 ASD (which are supposedly the same disorder). Moreover, at least 32% of people who have autism have ADHD, and vice versa (think AuDHD). ADHD and ASD are often hard to differentiate, especially when heavily relying on parent/teacher/self-report questionnaires and even the “gold standard” ADOS-2. As both seem to be inherited via similar genes, again begging the question, are they truly two separate disorders, and how does this affect the autism spectrum as a whole? Further, recently, Dame Uta Frith, a professor of cognitive development, stated that there is “no longer a common denominator and the spectrum has stretched so far it is no longer useful as a medical diagnosis.” This becomes controversial, because at present, there are no consistently agreed upon biomarkers, blood tests, genetic screenings etc (though this may soon change).
Aside from autism itself, a targeted autism or ADHD evaluation may also miss potential underlying medical explanations, such as PANS or PANDAS (medical conditions resulting from viral and bacterial infections that result in inflammation in the brain). PANDAS/PANS symptoms can include OCD, tics, separation anxiety, inattention, tics and sometimes neurological abnormalities such as or “brain fog.” These symptom were observed to have an abrupt onset and variable/episodic course, including “flare ups”, which can make assessment and identification even more challenging. On neuropsychological assessment, many PANDAS children had similar cognitive profiles as compared to those with ADHD, ASD, OCD, or anxiety (e.g., reduced visuospatial memory, reduced inhibitory control, reduced fine motor speed, and reduced sustained attention). Importantly, there were no clear differentiation in OCD, ASD, or tic disorders for those with and without PANDAS and no notable biomarkers or distinct clinical characteristics.
Another danger of conducting a targeted evaluation is the the potential of either misdiagnosing a child with autism when social/pragmatic communication disorder (SCD) and/or ADHD may be more appropriate. Similarly, a targeted ADHD evaluation could also miss (amongst other disorders). Importantly, SCD cannot be diagnosed if ASD is diagnosed, meaning they are mutually exclusive; however, ADHD can be diagnosed alongside SCD. One study found that up to 82% of children with ADHD experience some form of communication impairment that affected social development, further highlighting the strong overlap between ADHD and SCD. Others have expressed concern with the current classification system in terms of resolving this “grey area”, for which a child may have clear social communication deficits but minimal to no mild restricted/repetitive behaviors (e.g., perhaps tied to ADHD or anxiety), which may disqualify them for an autism diagnosis. Moreover, there are no formal measures that are specific to SCD, which further complicates diagnoses when specific questionnaires for ADHD and ASD are administered. Often times, during a comprehensive evaluation, clinical interviews are administered when autism is part of the referral question (e.g., ADI-2, DISCO, 3di), in addition to conceptualization of the overall neurocognitive profile, but this is not typical in a targeted evaluation. Moreover, many targeted evaluations will fail to look more closely at higher-level or pragmatic language skills, (e.g, CELF-5-META, SLDT-E:NU, TOPL-2) to pick up on more subtle pragmatic deficits.
Gifted children experience asynchronous development between academic and social-emotional/executive functioning skills, which may look like: questioning things often, focusing on big picture vs. details, a tendency toward fixation and hyper-focusing, and thinking differently and learning quickly. This type of profile can also mimic ASD, ADHD, and other neurodevelopment or anxiety disorders. From an emotional and behavioral perspective, often, the higher the IQ, the more the child struggles socially and emotionally, as they tend to experience anxiety and other emotions more intensely. Many will prefer interactions with older children and more often engage in solitary play, and often have different values, interests and feelings. From an academic perspective, it is typical for gifted children to demonstrate boredom and hyperactivity in the classroom. As a result, a targeted ADHD or ASD evaluation may miss aspects of intellect that are critical to diagnosis, and may best explain behavioral presentation. Similarly, children who are twice-exceptional may be under diagnosed in terms of ADHD, ASD, or learning disabilities, because of their high intellect and academic performance. Unfortunately, they can also be incorrectly labeled as ADHD due to reports of inattention, restlessness, and emotional reactivity, when those behaviors are actually a result of boredom or misfit in classroom placement, secondary to giftedness and emotional immaturity. On the other hand, maybe a child is mistakenly diagnosed with autism as a result of being labeled as “weird” and “antisocial”, and assumed to not understand what is going on around him, all the while he is processing information and understanding at a level significantly higher than many of the peers around him.
Unfortunately, targeted evaluations also lend themselves towards both confirmation and cultural bias more than comprehensive evaluations. Clinicians will sometimes form a first impression of a patient’s condition and then ask questions that confirm- but not refute- their impression (i.e., confirmation bias). In fact, one study found as low as 30% reliability in agreement of diagnosis between psychiatrists. In another study, it was found that when clinicians were asked to diagnose ADHD based on clinical vignettes and gender, they diagnosed the males twice as much, despite equivalent rates in the vignettes. In another study, a total of 13% of psychiatrists demonstrated confirmation bias when searching for new information after having made a preliminary diagnosis, resulting in inaccurate diagnoses for challenging cases. And lastly, in a randomized study of 475 clinicians who assigned diagnoses based on vignettes without additional assessment, there were several false-positive diagnoses of MDD, GAD, and borderline personality disorder. So how did this happen? Sometimes the clinician fails to gather sufficient information to rule out an alternative diagnosis, or challenge whether the former clinician’s original diagnosis was correct. Moreover, when parents or clients “cherry-pick” symptoms and conclude their problems are an indication of a certain disorder, they may disregard the importance of a full diagnostic process with other potential explanations. Information regarding the patient’s behavioral symptoms often rely on subjective reporting and sometimes questionable long-term memory, as well as having their own expectations of what appropriate/typical behavior is.
Importantly, though seemingly improving in the last decade, research has consistently showed that people of color are less likely to be diagnosed accurately. Black children are more likely to be misdiagnosed as ODD, ADHD, intellectually disabled, or with a mood disorder, rather than an appropriate diagnosis autism. Per the research, the most under-diagnosed population for autism is the Latinx community, especially for higher-functioning individuals. Further, there are significant cultural variations in what constitutes mental illness and what social communication and social conventions typically look like. Some cultures differ in what they consider to be age-appropriate. Procedures for screening autism may not adequately capture the experiences of parents from diverse backgrounds because they are very reliant on parent report, who may report more behavioral symptoms (sleep difficulties, aggression, language delay) vs developmental delays (missed milestones, eye contact, social motivation), resulting in focus on mental illness or learning disorder labels, as opposed to consideration of a developmental disorder. Studies have also shown that eating disorders and PTSD are under diagnosed in black teens as compared to white and hispanic teens. In addition, many evaluations do not use culturally relevant activities and/or questions on checklists/interviews, thus affecting validity of assessments. To further complicate things, many clinicians unfortunately can display covert racism or a lack of understanding, especially when dealing with a limited data set or when conducting a virtual evaluation.
Because, as mentioned, there is significant overlap in symptoms across disorders, this may result in a high rate of misdiagnosis and/or lack of diagnosis. Because of both under and misdiagnosis, there is lack of clarity and specificity in research pertaining to treatment for such diagnoses, often resulting in poor or tarnished data, ultimately affecting treatment outcomes. Therefore, intervention often comes too late, is not comprehensive, or inefficient. Some clinicians may lack expertise or niche experience with the elements of other disorders, which can also result in misdiagnosis or missed diagnosis. One study showed that in children, 18% of anxiety disorder diagnoses were missed by clinicians, while in another study, 29% of those who were diagnosed with depression were ultimately diagnosed with anxiety after formal structured interview. Moreover, another study revealed that those with a prior diagnosis of ADHD are less likely to receive an accurate autism diagnosis. It is also suspected that autism assessment has resulted in over diagnosis in boys and under diagnosis in girls, though this has been inconsistent in the research. On the other hand, one study found that a significantly larger portion of children who were (incorrectly) diagnosed with ADHD, were of the youngest in the grade, and therefore demonstrated significantly more immature social-emotional skills as compared to the older children in the class, despite similar neurocognitive presentation.
Another danger of targeted assessments lies in the fact that there is poor validity and reliability amongst clinicians when diagnosing, since many of the “definitions” of disorders are subjective and underlying constructs are poorly defined. Several diagnoses have the same subjective label they intend to define as one of the criteria (e.g., one of the symptoms required for a diagnosis of depression is “feeling depressed”). Many mental health tools (assessments/questionnaires) are heterogenous and overlapping, leading to ambiguity and poor standardization/validity/reliability of diagnoses. In fact, one study showed that, merely changing the threshold of clinical significance within a rating scale by one severity point, changed clinical prevalence by 26%. Targeted assessments typically take place in one setting as opposed to obtaining data from work/school/home/sports etc., which can also skew the diagnostic impression.
In closing, the potential danger here is that many times, through no fault of their own, parents or individuals end up bypassing a more involved comprehensive diagnostic evaluation (likely because of cost, waiting lists, insurance or geographical factors, etc.), for a briefer appointment, where they spend much less time with the doctor, where less historical or medical information is gathered from other providers. But, for some families or young adults who don’t have thousands to spend, this may be their best and only option. Often times, it feels like we have it backward. We diagnose based on behavioral symptoms- instead of identifying the reason for said symptoms. Think about it. When your child wakes up with a fever and a cough, you generally take them to doctor to get a diagnosis or understand WHY they have a fever, right? The doctor will typically run tests to see what underlying virus/bacteria/systemic issue etc. is causing the fever. Sure, you can treat the symptom of fever itself; however, this will not ultimately resolve the illness or prevent the fever from returning. My feeling is that we often neglect the cause or etiology of these “disorders” and tend to focus on the particular cluster of symptoms someone experiences, which may vary based upon how someone’s unique mind and body reacts to stress or trauma. The problem is that this can actually work against the benefit of a diagnosis, because there isn’t a true shared cause of the “illness.” For example, if every person with “anxiety” or “autism” received the exact same treatment (e.g., ABA, antipsychotics, mindfulness etc.), that approach could actually backfire. Additionally, a big concern is that if medications or other interventions are recommended based upon a biased or incorrect diagnosis, and the person does not receive appropriate intervention, this can be detrimental to their health and prognosis, as delayed diagnosis has been linked to increased relapse and hospitalizations and delayed remission.
In reality, there are often only very slight differences between those considered “normal” or “typical” and those considered “atypical” or “ill.” Even quantitatively, there is a very blurred line statistically between “normal” and “disordered.” And what if the diagnosis that best describes the child doesn’t exist in the DSM, like sensory processing disorder, pathological demand avoidance, PANDAS, or gifted intellect? Or what if the best “diagnosis” available really doesn’t seem to fit the child’s presentation? One of my favorite quotes is from Anna Mehler Papery, who put it so eloquently in her book Hello, I Want to Die, Please Fix Me, “The DSM’s authors boil down diagnosis of mental illness to something resembling an online quiz of: Which Disney Princess Mental Disorder Are You?” As such, a detailed developmental/social history and in-depth behavioral observations are critical for this diagnosis to be accurate. It can be challenging for clinicians to diagnose based upon similar profiles, without really understanding the quality, frequency, and severity of deficits or concerns in the context of the larger neuropsychological profile.