Diagnostic preferences include discussion of etiology for adults with cerebral palsy and their caregivers.

Abstract

Aim:

To determine the views of people with cerebral palsy (CP) and their caregivers regarding carrying a CP diagnosis, an etiologic diagnosis, or both diagnoses together.

Methods:

We surveyed people with CP and their caregivers across two registries querying their views on carrying a CP diagnosis, one type of etiologic diagnosis (specifically, a genetic diagnosis), or both. Open-ended responses were analyzed using a conventional content analysis approach.

Results:

Of 197 respondents (108 adults with CP and 89 caregivers), the majority (75%) valued knowing the cause of their CP. Of those with a diagnostic preference, the majority preferred carrying both CP and etiologic diagnoses together (68%). When compared to carrying an etiologic diagnosis alone, significantly more respondents felt a CP diagnosis helped anticipate symptom evolution (84% vs 54%), explain symptoms to others (86% vs 48%), access services (86% vs 48%), and join support communities (78% vs 50%) (p<0.01, Chi-square).

Discussion:

The majority of CP community members surveyed want to know the cause of their CP and would prefer carrying both CP and etiologic diagnoses together. Clinical practice should evolve to meet these community needs.

Introduction

Cerebral palsy (CP) is a non-progressive motor disability attributed to a disturbance in the developing fetal or infant brain.¹ It has long been established as a clinical diagnosis, independent of its etiology. Other neurodevelopmental disorders, like autism² and epilepsy^3,4, are also identified based on a clinical description but are typically provided together with etiologic diagnoses (e.g. autism associated with tuberous sclerosis complex⁵, or epilepsy due to a traumatic brain injury). Therefore, though autism and epilepsy are both clinical diagnoses, description of their associated etiologies is part of their diagnostic frameworks, without any consideration of removal of the autism or epilepsy diagnostic labels once their specific etiologies are identified.^2–4

However, practitioners tend to view CP diagnosis differently. That is, contrary to existing consensus statements^1,6, practitioners still impose etiologic restrictions when providing a CP diagnosis. This is particularly true for genetic etiologies: 40% of physicians would not provide a CP diagnosis in the setting of a genetic etiology.⁷ There were many cited reasons for this view. Some physicians noted that the term “cerebral palsy” was not specific and should be replaced with a more specific diagnosis when possible, with responses noting “cerebral palsy is a bad term” or that a CP diagnosis “does not allow for a deep understanding”.⁷ This may be because CP etiologies are multifactorial and heterogeneous (including but not limited to structural brain malformations, brain injury, and genetic etiologies) and it may be difficult to identify all causative factors in some people with CP. Practitioners may be wary of discussing this etiologic complexity or uncertainty with families.

Characterizing the divergent views of physicians on CP diagnosis is valuable. However, the effect of their diagnostic practices on people with CP and their families deserves equal, if not greater, consideration. Therefore, we sought to characterize the views of people with CP and their caregivers regarding CP diagnosis. We focused on CP community views with regards to genetic etiologies, primarily because provision of a CP diagnosis in the context of a genetic disorder has been associated with the greatest diagnostic variability amongst practitioners⁷ and because the role of genetic testing in CP is expanding rapidly over time.⁹ We hypothesized that the CP community would agree with existing consensus definitions that CP should be diagnosed just as all neurodevelopmental disorders are diagnosed: as a clinical diagnosis given together with an etiologic diagnosis.

Methods

This cross-sectional survey study received a human subjects research exemption from the Washington University Institutional Review Board (IRB Identification Number: 201908233).

Participant recruitment

Survey respondents were adults with CP (older than 18 years) or caregivers of a person with CP (of any age) who were responding on behalf of the person with CP they were caring for. There were no a priori restrictions to participation with regards to CP etiology or timing of the acquired injury etiology of CP. They were recruited via two organizations: the Cerebral Palsy Research Network (CPRN) and the Simons Searchlight Foundation.

The CPRN Community registry gathers data directly from people with CP and their caregivers via REDCap based surveys distributed to the community through MyCP.org.¹⁰ People with CP and their caregivers sign-up to participate through the MyCP.org portal and are presented with surveys that are relevant to them. The registrant’s CP diagnosis is ascertained by a standardized medical history intake form. The survey for this study was integrated into this portal.

The Simons Searchlight registry is an international online research registry for individuals with a genetic diagnosis associated with neurodevelopmental disorders ascertained by clinical or research testing.¹¹ Caregivers enrolled in Simons Searchlight were invited by email to complete the study if their dependent had a diagnosis of CP as reported in a standardized medical history interview. People included in the Simons Searchlight registry had at least one of the following: 16p11.2 deletion, 1q21.1 deletion copy number variations, or mutations in ADNP, ASXL3, CHAMP1, CSNK2A1, CTNNB1, DYRK1A, GRIN2B, HIVEP2, PPP2R1A, PPP2R5D, PCHD1, SCN2A, STXBP1 or SYNGAP1.

Registry enrollee characteristics

Registry enrollee characteristics were derived from existing registry data including age, Gross Motor Function Classification Scale (GMFCS) for the CPRN registry, and Vineland Adaptive Behavior Scale-II scores for the Simons Searchlight registry. The GMFCS is a five-level classification system that describes the self-initiated gross motor function of people with CP.¹² GMFCS level was self-reported in the CPRN registry. The Vineland Adaptive Behavior Scale-II measures adaptive behavior relative to established norms across the lifespan in five domains.¹³ The parent/caregiver rating forms for the motor skills domain were used form the Simons Searchlight registry. These measures were assessed to ensure that the respondent age and functional demographics were representative of the entire sampled population (i.e. all registry enrollees who received a survey link).

Based on CPRN and Simons Searchlight registry enrollment numbers, a total of 290 adults with CP (52%) and 270 caregivers of people with CP (48%) were contacted (N=560, 500 from the CPRN registry and 60 from the Simons Searchlight registry). The average age of the people with CP enrolled in these registries (either enrolling themselves as an adult or having been enrolled by a caregiver) was 26.5 years of age (standard deviation 17.9 years). In those for whom GMFCS data was available (n=500 from the CPRN registry), 305 (61%) were at GMFCS level I-II, which correspond to the highest degrees of gross motor ability.¹² In those for whom Vineland data was available (children under 7, n=28 from the Simons Searchlight registry), the average Motor Skills Doman Standard Score was 55.7 (95% CI 49.8–61.6) which corresponds to the “Low” range according to Vineland norms.¹³

Survey development

The survey was administered via REDCap between 12/1/2019 and 5/31/2020. The survey queried basic demographic information (age of the person with CP, who was responding – i.e. caregiver vs. person with CP, geographical location), known patterns of brain injury or potential CP etiologies, co-existing conditions, and questions about functional status across five domains adapted from validated functional classifications systems for CP.^14–18 Additionally, respondents were asked multiple choice questions about their views on knowing the cause of their CP, on carrying a CP and/or genetic diagnosis, and on any possible perceived benefits of carrying a CP or genetic diagnosis. These questions detailed in the survey instrument available as Supplementary Methods.

The survey was developed by all authors iteratively including input from a Simons Searchlight clinical psychologist with expertise in survey and protocol design (LGS), a neonatal neurologist with expertise in patient experience surveys (ML), and two CPRN leaders and community advocates and who are parents of children with CP (MS and PG).

Qualitative analysis

In addition to discrete response items, respondents were asked to explain why they might prefer a genetic diagnosis alone, a CP diagnosis alone, or both diagnoses, and were prompted: “Please explain why you chose this response regarding these diagnosis options”. These open-ended responses were analyzed using a conventional content analysis approach.¹⁹ In the absence of an established framework for analyzing CP diagnostic preferences, we created and refined a novel codebook to characterize response content. Two investigators (ML, BA) independently coded all responses. Discrepancies were resolved in consensus.

Statistical analysis

Statistical analyses were undertaken using SPSS (IBM, Armonk, NY). Significance levels were set a priori at p<0.05. Fisher’s exact tests with Bonferroni corrections for multiple comparisons were used to determine which, if any, of the four categorical respondent characteristics might be associated with respondent diagnostic preferences (i.e. significance level was set at p<0.05/4 or p<0.0125). Chi-square tests with Bonferroni corrections were used to determine whether there were differences in the perceived benefits of carrying a CP diagnosis versus a genetic diagnosis (i.e. significance level was set at p<0.05/5 or p<0.01 as comparisons were done across five potential perceived benefits). Chi-square and t-tests were used to compare proportions and ages as indicated.

Results

Respondent characteristics

Of the 560 people in the CPRN and Simons Searchlight registries who received the survey link, 197 people responded (35% responder rate). Of these respondents, 89 were caregivers responding on behalf of the people with CP they care for (45%) and 108 were adults with CP (55%). The vast majority lived in the United Sates (175/197, 89%). Of The mean age of the person with CP (who was either responding for themselves or being responded for by their caregiver) was 27.7 years (standard deviation 18.0 years), with 71/197 (36%) less than 18 years old. Over 50% of respondents were at functional classification levels I-II (high functional status) across all queried domains (gross motor, fine motor, communication, eating/drinking, and vision) (Table 1).

Table 1.

Respondent demographics.

Respondent demographic	N (%, out of 197 total respondents)
Age  <18 years old  18–60 years old  >60 years old  Missing data	71 (36%) 112 (57%) 14 (7%) -
Relationship to person with CP  Self (adult with CP)  Caregiver  Missing data	108 (55%) 89 (45%) -
Location  United States  Canada  Other  Missing data	175 (89%) 8 (4%) 14 (7%) -
Brain injury pattern or potential etiology (multiple could be selected)  Periventricular leukomalacia  Intraventricular hemorrhage  Hydrocephalus  Stroke  Hypoxic-ischemic encephalopathy  Genetic condition  Infection  None / I don’t know  Missing data	28 (14%) 26 (13%) 24 (12%) 21 (11%) 19 (10%) 26 (13%) 2 (1%) 64 (32%) -
Non-motor co-existing condition (multiple could be selected)  Epilepsy  Autism  ADHD  Other*  None  Missing data	44 (22%) 20 (10%) 18 (9%) 38 (19%) 77 (39%) -
GMFCS (Gross motor)  I  II  III  IV  V  Missing data	48 (24%) 70 (36%) 19 (10%) 32 (16%) 22 (11%) 6 (3%)
MACS (Fine motor)  I  II  III  IV  V  Missing data	72 (37%) 65 (33%) 13 (7%) 23 (12%) 13 (7%) 11 (6%)
CFCS (Communication)  I  II  III  IV  V  Missing data	97 (49%) 40 (20%) 22 (11%) 13 (7%) 11 (6%) 14 (7%)
EDACS (Eating and drinking)  I  I  III  IV  V  Missing data	107 (54%) 42 (21%) 17 (9%) 15 (8%) 5 (3%) 11 (6%)
VFCS (Vision)  I  II  III  IV  V  Missing data	132 (67%) 28 (14%) 16 (8%) 9 (5%) 1 (1%) 11 (6%)

^*Other non-motor co-existing conditions included: central/cerebral visual impairment, sensory processing disorder, hearing loss, anxiety, bipolar disorder, asthma, hypertension, gastroparesis and cyclical vomiting. GMFCS – Gross motor function classification system, MACS - Manual ability classification system, CFCS – Communication function classification system, EDACS – Eating and drinking classification system, VFCS – Visual function classification system.

There was no significant difference between respondents and the entire registry population who received the survey with regards to age (p=0.35, t-test) or the proportion of caregivers versus adults with CP (Chi-square 0.54, df 1, p=0.46). Noting that CP is primarily defined by the presence of a persistent non-progressive motor disability,¹ there was also no significant difference between respondents and the entire registry population with regards to GMFCS level (Chi-square 0.035, df=1, p=0.85, comparing proportions of those at GMFCS levels I-II vs. III-V). Functional information in the fine motor, communication, eating/drinking, and visual domains was queried for the respondents but not consistently available for the entire sample population.

The average age of respondents at the time they were diagnosed with CP was 1.55 years old (standard deviation 0.97 years). Almost half of respondents noted that they had non-motor co-existing conditions accompanying CP (90/197, 46%), the most common of which was epilepsy (44/197, 22%). Twenty-six respondents stated that they carried a genetic diagnosis; of these individuals, 20 were enrolled in the Simons Searchlight registry and 6 were enrolled in the CPRN registry. 21 of these 26 individuals described the affected genes/chromosomal regions including 7 with 16p11.2 deletions²⁰, 4 with ADNP mutations²¹, 4 with STXBP1 mutations²², and 1 respondent each with mutations in PPP2R5D²³, PPP2R1A²⁴, ADCY5²⁵, SCN2A²⁶, GRIN2B²⁷, and 1q21.1 deletion²⁸. Of note, all of these disorders have been associated with motor disability, including predominant hypotonia with or without a combination of other motor symptoms like spasticity, dystonia, ataxia, and tremor.

The majority of respondents (117/197, 59%) were able to identify a brain injury pattern or potential etiologic mechanism associated with their CP, and many indicated more than one brain injury pattern or potential etiologic mechanism (29/197, 15%) (Table 1). However, only 48% (90/188) noted that a physician had told them the cause of their CP. Dissecting this data further, of those who identified a brain injury pattern or potential etiologic mechanism associated with their CP, only 64% (75/117) noted that a medical practitioner had told them the cause of their CP. This is particularly notable given that the majority of respondents (147/197, 75%) valued knowing the cause of their CP.

Respondent views on carrying a CP diagnosis and/or a genetic diagnosis

Caregivers and people with CP were also asked questions regarding whether they would prefer to carry a CP diagnosis alone, an etiologic diagnosis alone (namely, a genetic diagnosis, given that genetic CP etiologies have resulted in the most diagnostic variability amongst practitioners), or carry both diagnoses together, if applicable. Most respondents indicated that they would prefer to carry a CP diagnosis (108/195, 55%), followed by “unsure” or no preference (80/195, 41%). Only 4% (7/195) would prefer to carry a genetic diagnosis alone. Of those who wanted to carry a CP diagnosis, a high number specified that they would prefer to carry the CP diagnosis together with a genetic diagnosis (78/108, 72%). Of those who had any preference regarding their diagnosis, 68% (78/115) preferred carrying both a CP diagnosis and a genetic diagnosis (Table 2).

Table 2.

Respondent diagnostic preferences.

Respondent characteristics	Respondent diagnostic preference				Fisher’s exact test (value, p)
	Etiologic diagnosis only (7/195, 4%)	CP diagnosis only (30/195, 15%)	Both diagnoses (78/195, 40%)	Unsure or ambivalent (80/195, 41%)
	N, %	N, %	N, %	N, %
Respondent identity   Adult (110/195, 56%)   Caregiver (85/195, 44%)   Missing data (2/197, 2%)	0/110 (0%) 7/85 (8%)	19/110 (17%) 11/85 (13%)	42/110 (38%) 36/85 (42%)	49/110 (45%) 31/85 (36%)	10.0, p=0.017
Independently ambulatory?   Yes (136/190, 72%)   No (54/190, 28%)   Missing data (10/190, 5%)	4/136 (3%) 1/54 (2%)	24/136 (18%) 6/54 (11%)	51/136 (38%) 26/54 (48%)	57/136 (42%) 21/54 (39%)	2.2, p=0.55
Non-motor co-morbidities?   Yes (91/195, 47%)   No (104/195, 53%)   Missing data (2/197, 2%)	6/91 (7%) 1/104 (1%)	13/91 (14%) 17/104 (16%)	42/91 (46%) 36/104 (35%)	30/91 (33%) 50/104 (48%)	8.6, p=0.033
Etiology known?   Yes, genetic (25/195, 13%)   Yes, acquired/not genetic (91/197, 47%)   No (79/195, 40%)   Missing data (2/197, 2%)	3/25 (12%) 2/91 (2%) 2/79 (1%)	1/25 (4%) 15/91 (16%) 14/195 (18%)	16/25 (64%) 40/91 (44%) 22/79 (28%)	5/25 (20%) 34/91 (37%) 41/79 (52%)	18.9, p=0.003*

Fisher’s exact tests were used to determine which respondent characteristics were associated with respondent diagnostic preferences (given respondent numbers less than 5 in some groups). Accounting for a Bonferroni correction for four different comparisons (respondent characteristics), the significance level was set at

^*p<0.0125 (0.05/4).

Significantly more respondents felt that carrying a CP diagnosis provided certain advantages over carrying a genetic diagnosis: a high proportion indicated that a CP diagnosis provided a way to anticipate the evolution of their symptoms (84% vs 54%), explain their symptoms to others (86% vs 48%), gain access to services (86% vs 48%), become part of a community of people with similar symptoms (78% vs 50%), and understand the cause of their symptoms (71% vs 53%) (p<0.005, Chi square with Bonferroni correction for multiple comparisons) (Table 3).

Table 3.

Perceived benefits of a CP diagnosis or genetic diagnosis.

	CP diagnosis	Genetic diagnosis	Chi-square (value, p)
Helps anticipate evolution of symptoms?	163/195 (84%)	105/195 (54%)	40.1, p<0.001*
Helps explain symptoms to others?	168/195 (86%)	93/195 (48%)	65.2, p<0.001*
Helps gain access to services?	168/195 (86%)	93/195 (48%)	65.2, p<0.001*
Helps become part of a community of people with similar symptoms?	152/195 (78%)	97/195 (50%)	33.6, p<0.001*
Helps understand the cause of their symptoms?	138/195 (71%)	104/195 (53%)	12.6, p<0.001*

Chi-square p-values are shown taking into consideration a Bonferroni correction for multiple comparisons (df=1). Accounting for a Bonferroni correction for five different comparisons (potential perceived benefits), the significance level was set at

^*p<0.01 (0.05/4). Missing data: 2/197 (1%)

We sought to determine whether there were certain respondent characteristics that were significantly associated with respondent diagnostic preferences. We considered the following: respondent identity (caregiver or adult with CP), gross motor functional status (independently ambulatory or not), non-motor comorbidities (present or not), and knowledge of their CP etiologies (genetic etiology, non-genetic/acquired etiology, or no etiology known). Correcting for multiple comparisons, only a personal knowledge of their potential CP etiologies was significantly associated with diagnostic preferences (Fisher’s exact value 18.9, p=0.003). Those who carried a CP diagnosis together with an etiologic diagnosis most commonly expressed a preference for keeping both diagnoses (56/116, 48%), regardless of whether their CP etiology was genetic (16/25, 64%) or not (40/91, 44%). In contrast, respondents who carried a CP diagnosis without knowledge of their CP etiology were most often unsure or ambivalent regarding their diagnostic preferences (41/79, 52%) (Table 2).

Respondent rationale for their diagnostic preferences

Of the 197 respondents, 147 provided free text explanations for the rationale underlying their diagnostic preferences regarding carrying a CP diagnosis alone, a genetic diagnosis, or both diagnoses together (Table 4).

Table 4.

Self-described respondent rationale for their diagnostic preferences as determined using conventional content analysis.

Respondent rationale (147/195, 75%)	Respondent diagnostic preference
	Etiologic diagnosis only (7/195, 4%)	CP diagnosis only (30/195, 15%)	Both diagnoses (78/195, 40%)	Unsure or ambivalent (80/195, 41%)
More knowledge (76/195, 39%)	2/76 (3%)	1/76 (1%)	53/76 (70%)	20/76 (26%)
Treatment/therapies (27/195, 14%)	0/27 (0%)	2/27 (7%)	16/27 (59%)	9/27 (33%)
More resources (24/195, 12%)	0/24 (0%)	2/24 (8%)	20/24 (83%)	2/24 (8%)
Does not matter / would not affect me (25/195, 13%)	0/15 (0%)	0/15 (0%)	3/15 (20%)	12/15 (80%)
What we were told / have always done (20/195, 10%)	2/20 (10%)	8/20 (40%)	2/20 (10%)	8/20 (40%)
Most accurate explanation of symptoms (12/195, 6%)	2/12 (17%)	0/12 (0%)	8/12 (67%)	2/12 (17%)
Help others understand (10/195, 5%)	0/10 (0%)	0/10 (0%)	9/10 (90%)	1/10 (10%)
Family planning (10/195, 5%)	0/10 (0%)	0/1 (10%)	7/10 (70%)	2/10 (20%)
Concern about the implications of a genetic diagnosis (9/195, 5%)	0/9 (0%)	8/9 (89%)	0/10 (0%)	1/9 (11%)
Uncertainty (7/195, 4%)	0/7 (0%)	0/7 (0%)	0/7 (0%)	7/7 (7%)
Assuaging guilt (3/195, 2%)	0/3 (0%)	0/3 (0%)	3/3 (100%)	0/3 (0%)
Change is stressful (1/195, <1%)	0/1 (0%)	1/1 (100%)	0/1 (0%)	0/1 (0%)

Missing data: 2/197 (2%) on their diagnostic preferences; 50/197 (25%) on their rationale for their diagnostic preferences

Commonly cited reasons underlying respondent diagnostic preferences included the need for more knowledge (76/195, 39%), the importance of treatment (27/195, 14%), and the need for more resources or services (14/195, 12%). Most respondents who cited these benefits also preferred to carry both a genetic diagnosis and CP diagnosis together (n=89/127, 70%). Some participants described how both diagnoses would bring clarity, including an adult with CP who preferred carrying both a CP and genetic diagnosis together because: “I feel like that would make the picture whole.” For others, the potential for more than one diagnosis offered access to community. As shared by one parent: “When my child gets older, I would want him to be able to connect with as many people as possible with a shared identity.” Many respondents highlighted the complementary benefits of each diagnosis, for example:

“The genetic diagnosis is what caused the cerebral palsy, which is important information, but it does not describe the person’s condition. Not everyone with the same genetic diagnosis [has] the same symptoms. Also, the genetic diagnosis is rare and not many people, even medical professionals, may be familiar with it, so cerebral palsy is a general term that most people understand. Also a cerebral palsy [diagnosis] will allow the patient to get more resources such as funding for equipment and therapy.”

– Caregiver of an adult with CP

Some respondents shared how a genetic diagnosis might impact family planning; a few respondents shared how a diagnosis might assuage maternal guilt: “For some parents, especially mothers, it would be comforting to know that there wasn’t any blame on their part - living with any guilt is, or can be extremely difficult.” A minority of respondents reported that a diagnosis would not affect their lives (25/195, 13%) or reiterated what they had already been told about their diagnosis (20/195, 10%).

Discussion

Our results show that 75% of the CP community (either people with CP or caregivers speaking on their behalf) values knowing the cause of their CP. However, only 48% had been explicitly told the cause of their CP by a medical practitioner. Of those with a diagnostic preference, 68% favored receiving both a diagnosis that describes their symptoms (i.e. CP) and a diagnosis that describes the cause of those symptoms. Therefore, we demonstrate support in the CP community for giving a CP diagnosis in addition to an etiologic diagnosis. This practice is aligned with diagnostic conventions for other neurologic disorders like autism² and epilepsy³ that are clinical diagnoses given together with their etiologic diagnoses.

Though we have previously shown that many physicians would not diagnose CP in a person with a genetic etiology for their CP symptoms,⁷ respondents noted that a CP diagnosis provides clarity, services, and support more often than does a genetic diagnosis alone. Therefore, in addition to being an accurate clinical description of the person’s phenotype, a CP diagnosis can serve as an important tool for a person’s understanding of their symptoms and can provide access to the larger CP community, which has been shown to improve the well-being of caregivers of people with CP.²⁹ Furthermore, people who carried a genetic diagnosis were significantly more likely to want to also carry a CP diagnosis, if applicable, obviating the potential practitioner concern that a single “more specific” diagnosis may be preferable for the affected person and their family. These results underscore the importance of considering the broader social environment in which people with CP live when providing medical diagnoses.

Clinicians should be explicit when discussing CP etiology. It is notable that 59% of respondents had been told that they had a brain injury pattern (e.g. periventricular leukomalacia or intraventricular hemorrhage) or potential etiologic mechanism (e.g. hypoxic-ischemic encephalopathy or infection) associated with their CP, but only 48% stated that a medical practitioner had told them the cause of their CP. More specifically, only 64% of people with knowledge of a brain injury pattern or potential etiologic mechanism associated with their CP had explicitly been told by a medical practitioner that any of those factors caused their CP. This suggests a possible discrepancy between practitioners sharing information on potential CP etiologies or risk factors and practitioners explicitly communicating that these factors could have caused CP. This is a potential practice gap that we hope these results will compel practitioners to address.

Clinicians may also not be adequately conveying the meaning and potential impact of a CP diagnosis or an etiologic diagnosis during clinical care. A large proportion of respondents (41%) were unable to articulate specific diagnostic preferences, noting that they were either unsure of which diagnostic option would make the most sense for them or were willing to accept any of the available diagnostic options. Clinicians should address this uncertainty with people they are evaluating for CP. Conferring a CP diagnosis early and accurately can be critical for prompt initiation of potentially outcome modifying therapies and treatments.³⁰ An etiologic diagnosis can inform treatment options, family planning, prognostication, and anticipation and screening for co-morbidities.^6,31 Therefore, clinicians should discuss possible CP etiologies during clinic visits openly and explicitly, including addressing their concerns regarding what these etiologic diagnoses may portend for themselves and their families.

A small number of respondents discussed feelings of caregiver guilt regarding the presumed CP etiology that might be assuaged if a genetic etiology for CP could be identified. Though this was a minor theme in terms of frequency, it is an important one. It is possible that explicit discussion with caregivers regarding CP etiologies can open conversations to address caregiver feelings of likely unfounded guilt. As one respondent put it, “living with any guilt is, or can be extremely difficult.”

The difficulty of identifying CP etiologies cannot be overstated. For many, CP could be the result of multifactorial etiologies (as indicated by 15% of our respondents who noted multiple brain injury patterns and potential etiologic mechanisms associated with their CP). For others, the etiologies may be unclear. Practitioners may be reticent to discuss this etiologic complexity and uncertainty with those affected by CP. However, there is growing data that open communication on these issues is both expected and preferred by the CP community.^8,32 Our results further support this growing body of literature.

Limitations

This survey required participants to ascribe a value to an etiologic diagnosis for which they may have little knowledge (i.e. a hypothetical genetic diagnosis), which is a difficult task. There is also respondent bias toward staying with their current diagnosis, as explicitly noted by 10% of respondents (Table 4). These challenges may have contributed to the relatively high proportion of unsure and neutral responses to the questions. However, noting that many people without an etiologic diagnosis would still be interested in carrying both a CP diagnosis and an etiologic diagnosis together, it is possible that these biases were at least partially overcome within the hypothetical context of this survey. Future work should determine whether these results can be independently replicated.

We focused on genetic etiologies of CP in our survey. Therefore, it is unclear whether the CP community would have different views regarding carrying an etiologic diagnosis that was not genetic. Noting that practitioners are more weary to give people a CP diagnosis in the setting of a genetic etiology⁷, and noting that 5% of respondents indicated trepidation regarding the implications of a genetic diagnosis (Table 4), it is possible that community views may actually be more favorable regarding carrying a CP diagnosis together with a non-genetic etiologic diagnosis. This requires further study.

The majority (55%) of respondents were adults with CP who had been given their CP diagnoses in childhood. It is possible that these individuals underwent their diagnostic CP evaluations several years ago and have not had these evaluations revisited despite significant recent advancements in imaging and genetic testing techniques. Being temporally remote from their diagnostic evaluations likely influenced respondent preferences and could have affected their recollection of what diagnostic evaluations were done. Future work could determine whether medical practitioners re-initiating discussions of CP etiology for adults with CP would change their views on the value of carrying a CP diagnosis together with an etiologic diagnosis.

The respondents who noted that they had a genetic diagnosis were disproportionately recruited from the Simons Searchlight registry. Though this was by design to ensure that people with known genetic diagnoses were adequately represented in the sample population, it is also a relatively narrow representation of the diverse population of people with CP with genetic etiologies. Future research should address the views of people with known genetic etiologies of CP more thoroughly, across a larger number of respondents.

We did not query whether respondents who had received a CP diagnosis had predominantly hypotonia as the cause of their motor disability. There remains controversy in the field regarding whether a person with pure hypotonia as a cause of their motor disability can be classified as having CP or not.^7,33 Our data suggest that the overarching benefits of carrying a CP diagnosis enumerated by our respondents could warrant extension of a CP diagnosis to those with so-called “hypotonic CP”. However, this topic warrants further debate.

Recommendations for providing a CP diagnosis

These stakeholder-driven findings support long-standing CP diagnostic criteria and registry inclusion criteria stating that CP is a clinical diagnosis separate from an etiologic diagnosis.^1,6,33–36 Our CP diagnostic recommendations, therefore, stem not just from the results of this survey, but from the collective understanding of the CP community at large: clinicians and researchers together with people with CP and caregivers.

When making a CP diagnosis, we recommend that physicians additionally provide diagnoses describing the CP etiology or etiologies. These etiologies should be described in terms of two categories: 1) mechanisms/risk factors, and 2) brain malformation/injury patterns. Documentation examples could include:

1. Birth at 24 weeks gestation associated with periventricular leukomalacia;

2. Perinatal CMV infection resulting in periventricular calcifications and left frontal gray matter heterotopias;

3. Non-accidental trauma at 4 months old resulting in bifrontal hemorrhages and diffuse axonal injury

4. CTNNB1-associated genetic disorder without evidence of malformation or injury on brain MRI

5. Birth at 28 weeks gestation together with prenatally diagnosed L1CAM-associated congenital hydrocephalus. Term-equivalent brain MRI additionally showed periventricular leukomalacia.

It could be difficult to determine which brain imaging findings are most salient when considering the contribution to the CP phenotype. The Surveillance of Cerebral Palsy in Europe MRI Classification System could serve as a standardized way to categorize brain MRI findings and may help mitigate some of this uncertainty.³⁷

In the setting of multifactorial etiologies, all etiologic contributors to a CP phenotype should be described. Finally, if an etiology is suspected but not confirmed or if the etiologies of CP are uncertain, this uncertainty should be communicated clearly to the person with CP and their family.⁸ This discussion is likely to be both long and longitudinal as ongoing etiologic investigations inform an evolving understanding of an individual’s CP etiologies. Advances in both imaging and genetic testing modalities means that etiologic investigations also morph over time and, therefore, so does the understanding of the CP etiologies for any given person. Whole exome sequencing can report genetic variants of uncertain significance (VUS). The fact that such variants exist should also be communicated to the family, including the plans and timeline for whole exome re-analysis.³⁸ Documentation examples could include:

1. Birth at 37 weeks gestation with a benign pregnancy, delivery, and post-natal history. Brain MRI is unremarkable. A genetic etiology is suspected with chromosomal microarray unremarkable and whole exome sequencing in 2019 identifying 2 VUS. Plan for re-analysis in 2021–2022.

2. Adoption occurred at age 2 with birth history unknown, though the family was told that the birth occurred “too early”. There is bilateral periventricular leukomalacia on brain MRI. The CP etiologic risk factor is assumed to be prematurity, though this is not confirmed.

It is of note that the examples provided above may be sufficient for documentation in the medical chart, but that discussions of these issues with families require much more detail and nuance. The appropriate approach to take for these discussions should be a topic for future research. Counseling regarding the implications of extant and potential CP etiologies can be complex, and could require referral to a geneticist, genetic counselor, neurologist, and/or other subspecialist well-versed in the diagnostic evaluation of CP.

In sum, we suggest clinicians communicate CP and etiologic diagnoses guided by the principles below:

1. Explicitly discuss each etiology of CP together with the evidence that led to identification of that etiology, including birth history, imaging, genetic testing, family history, and motor phenotype. Any associated brain injury or patterns of brain malformation should also be discussed.

2. For people who have multiple contributing CP etiologies, clearly communicate all of them.

3. For people where the etiological contributors to the CP phenotype are unclear, disclose this uncertainty to the family and consider referral to a specialist for further etiologic evaluation and genetic testing if it has not already been done.^8,39

4. Discuss the implications of a CP diagnosis with families, including treatment, prognosis, screening for co-existing conditions, access to services, and access to other support systems.

5. Discuss the implications of each etiologic diagnosis with families, including any implications for family planning in addition to the above features.

Conclusions

Though many clinicians may believe that an etiologically-specific diagnosis is preferable to a broader phenotypically-based diagnosis like CP,⁷ people with CP and their caregivers generally prefer carrying a CP diagnosis together with the etiologic diagnosis (including a genetic diagnosis). Therefore, just as autism² or epilepsy³ diagnoses and their etiologies, genetic or otherwise, are not mutually exclusive,^40,41 these results demonstrate community support for conferring a CP diagnosis together with etiologic diagnoses. This approach is in line with existing CP diagnostic criteria and, moreover, reflects the majority views of people and families directly affected by CP.

What this study adds:

• Most CP community members value knowing their CP etiology

• Most community members with diagnostic preferences desire both etiologic and CP diagnoses

Data availability:

De-identified respondent quotes are available in Supplemental Results. Anonymized data will also be shared by request from any qualified investigator.