Basic symptoms (BS) in schizophrenia are defined as subtle and subclinical alterations experienced in various areas of mental life.1 Although the presentation of the BS varies between subjects, these subjective experiences englobe mostly neurocognitive alterations, affective alterations, and subclinical alterations of perception and thoughts.2

In the last decade, many research groups have worked on developing instruments that measure BS since their assessment could also help detect early schizophrenia.3 An example of those instruments is the Eppendorf Schizophrenia Inventory (ESI),4–6 which measures BS phenomena such as speech and attention, ideas of references, auditory uncertainty, and dispersion of perception. Initial results have shown the advantages of ESI for measuring BS, demonstrating superiority over the Frankfurt Discomfort Questionnaire (FCQ).5 Additionally, its application was uncomplicated for the participant, being quick to fill out and untroubled to answer.

The diversity of populations worldwide and cross-cultural and international research indicate a necessity for validated instruments for different groups of people and languages. In the case of the ESI, for instance, there are several translations published (e.g., Korean or Italian),6,7 except for the Spanish language, being the latter one of the most spoken languages, with about 491 million speakers worldwide.8 Additionally, in some emerging Spanish-speaking nations, such as Peru, where scientific production is progressively increasing, tools, such as the ESI, are required for schizophrenia research and clinical purposes. In the case of subjective symptoms of schizophrenia, an instrument such as the ESI will enable the physician to determine the individuals at high risk for developing psychosis. In addition to being unknown and necessary in Spanish-speaking countries, such as Peru, this information must be disseminated to the world scientific community since few instruments evaluate initial symptoms of schizophrenia in the Spanish language.

Therefore, the main aim of this study is to apply content validation to the first Spanish version of the ESI (ESI-Sp) using the Aiken methodology through expert judgment.

Brief Description of the Instrument – ESI: The Eppendorf Schizophrenia Inventory (ESI) was created by Mass et al. and initially written in German5 to evaluate schizophrenia symptoms, especially at the early onset of the psychotic disorder. This instrument consists of 40 items (response options: 4 – true, 3 – mostly true, 2 – sometimes is true, 1 – that is not true) and four different factors5: attention and speech impairment (AS, 10 items), ideas of reference (IR, 7 items), auditory uncertainty (AU, 8 items), and deviant perception (DP, 9 items). As “control” questions, the “Frankness” group (FR, 5 items) was included, questions that assess the willingness or ability of the subjects surveyed to admit minor personal flaws in a general way.5

Translation and cultural adaptation: The original version in German was translated into Spanish by one of the authors (BPP). Then, we invited four psychiatrists from Lima (Peru) with the following characteristics: ten years of clinical experience and active professors in clinical Psychiatry. Our main goal for these four experts was to suggest changes regarding the wording and readability without changing the content of the items. Afterward, one of the authors (WH) re-translated the adapted instrument. The re-translated version was sent to the original author of the ESI to compare and contrast with the original version in German, obtaining thus the approval to use. Table 1 shows the adapted items in Spanish.

Expert judgment and validation: After translation and cultural adaptation, we sent the translated version to different raters, a group of Peruvian psychiatrists with clinical experience. The selection of the raters was made by using the snowball method. Here, we introduced the translated version on an online free survey website (Google Forms) and distributed it using different social media or electronic mail addresses.

Ethical statements: This expert validation study was approved by the ethics committee of the Peruvian University “Cayetano Heredia”. The ethical procedures of the study were carried out according to the Helsinki Declaration. Finally, our study group received the translation rights of the ESI from the Hogrefe Editorial for research purposes.

Content validity using Aiken Methodology (Aiken's V): Content validity was calculated using the Aiken methodology (Aiken's V, V=0: total disagreement, V=1: total agreement, V=0.5: indecision), which evaluates the number of agreements between experts. To measure the agreement between judges, we formulated two questions for each item (Clarity: Do you consider this item clear? Representativeness: Do you think this item's representativeness is adequate?). Both questions were scored using a Likert scale of 5 categories.

Aiken's V9,10 was calculated from the judges’ information on both elements. Confidence intervals corresponding to 95% (CI95)9,11 were also calculated, which allowed estimating the population parameters for the value of V (VP). The null hypothesis was defined as the probability that the VP takes values equal to or less than 0.50. Therefore, to accept the null hypothesis, the CI95 must include the value of V≤0.5 within the range of values, giving the possibility that VP≤0.50. To calculate each CI95 for Aiken's V, we used the methodological proposal from Penfield and Giacobbi.9 Finally, the formulae for the Aiken methodology concerning content validation were calculated in Microsoft Excel 2019, following the recommendations published elsewhere.9,11

Translation and cultural adaptation: We did not find any items that needed alteration during translation, cultural adaptation, and retranslation. The contents of the questions remained stable with the ones of the original version. The final result of this phase is presented in Table 1.

Brief description of the judges: Initially, 32 judges were summoned to participate. However, two of them retired their participation for unknown reasons, not filling out the online survey for the ESI evaluation. Therefore, thirty judges (clinical psychiatrists; n=30; 47.07±13.62 years) completed the study and were included in the analysis. The majority of our participants were psychiatrists working in an institute (n=10, 33.33%), followed by psychiatrists working in a hospital (n=6, 20.00%), private practice (n=6, 20.00%), community centers (n=4, 13.33%), university (n=2, 6.67%), forensic institutions (n=1, 3.33%), and private clinic (n=1, 3.33%). Regarding the years of experience, the majority had “15 or more” years of experience (n=12, 40.00%), followed by “between 2 and 10” years of experience (n=11, 36.67%). Finally, a small group of colleagues indicated “between 11 and 15” years of experience (n=7, 23.33%). Concerning the field of expertise, most of the raters worked in adult psychiatry (n=20, 66.67%).

Content validity using Aiken's V: We performed the content validity analysis using Aiken's methodology. The analysis was performed for each item, the factors, and the whole instrument, as presented in Table 2. Concerning the 40 items of the ESI, the null hypothesis, defined in the methods sections, was rejected for all items and demonstrated the content validity of each item (Table 2). Similar results were found for the four factors of the ESI: AS (V=0.75; CI95 [0.67; 0.82]), IR (V=0.81; CI95 [0.74; 0.87]), AU (V=0.81; CI95 [0.73; 0.87]), and DP (V=0.81; CI95 [0.73; 0.87]). Finally, the total Aiken's V for the ESI also did not include the values of the null hypothesis in the confidence intervals (V=0.78; CI95 [0.70; 0.85]).

Our results suggested that the first ESI-Sp has adequate representativeness and a degree of clarity following Aiken's methodology. Furthermore, the wording and readability changes during translation, cultural transformation, and retranslation remained stable for all the items of the ESI-Sp, showing similarities with the literature recommendation concerning qualitative validation.12,13

Content validity analysis of the items individually and four factors through Aiken's V methodology showed significant results, consistent with the experts’ judgment that the first ESI-Sp is valid. However, given our study's design, we cannot wholly validate this instrument in people with schizophrenia onset or ultra-high-risk psychosis since the study for construct validation is missing. This qualitative expert validation study is the first step for an ESI validation project; we will proceed with a construct validation study with these results.

We have detected some strengths and limitations of our study and highlighted them. The primary strength of this short communication is the translation, cultural adaptation, and Expert judgment of the first Spanish version of this instrument. However, the main limitations of our study should be addressed. Firstly, even though the number of experts calculated for the sample size is enough, the sampling technique is non-probabilistic, increasing the risk of sample bias. Finally, the study design is a limitation of this study since it is impossible to make a statistical inference or, in this case, a construct validation based solely on the experts’ opinions.

In conclusion, our study suggests that the first ESI-Sp is a valid and representative instrument to determine subjective experiences in patients with a high risk of schizophrenia. The construct validation of the ESI-Sp is pending, which our research group will carry out after this first step.

The authors declare they have no conflict of interest.