Table 1.
Various categories of problems with the data in CTRI, comparison with the relevant fields of ClinicalTrials.gov where possible, and suggestions to improve CTRI in future
| No. | Category in paper | CTRI | ClinicalTrials.gov | Suggestion to improve CTRI in future |
|---|---|---|---|---|
| A. Lack of clarity in the classification of Type of Study | ||||
| 1 | 1. Type of Study |
We needed to consider two fields: 1. Under Type of Trial, there were four options, of which the registrant had to choose one from a drop-down menu: interventional, observational, BA/BE or postmarketing surveillance (PMS) 2. Under Type of Study, there were 18 fields, and any number could be chosen. In April 2018 we downloaded the data of all 12,673 trials. These trials fell into 1331 categories, which were rather confusing. Examples included: DrugSurgical/Anesthesia, DrugAyurveda,a DrugSiddha, Drug Preventive, DrugBiological, DrugMedical Device, and DrugOther |
There were two equivalent fields: 1. Study type had 3 options, Interventional studies, Observational studies (including Patient Registries), and Expanded Access studies 2. Intervention had 11 categories.b One category could be chosen more than once, and more than one category could also be chosen. However the listing of multiple interventions was discrete and unambiguous, as follows: Example 1: Drug: 1.5 mg estradiol and 2.5 mg nomegestrol acetate|Drug: 15 μg ethinylestradiol and 60 μg gestodene Example 2: Radiation: [C-11]|Drug: NOP-1A |
In order for registry users to rapidly understand the nature of a trial, registrants should classify each trial using a multi-layered menu of options, along the following lines: The first layer should be the system of medicine to which the trial belongs, and the options should be (1) allopathy (mainstream medicine) and (2) AYUSHa (alternative medicine). Within each system, the choices for Type of Trial should remain as they are. Within interventional trials the choices should be prevention, screening, treatment, education, and others. Under treatment, a distinction should be made between small molecules, biologics, herbal remedies, or a combination thereof. Within biologics, a distinction should be made between vaccine, stem cell therapy, biological, or a combination thereof. Further details would need to be worked out to enable each trial to be optimally and comprehensively classified The output format must be unambiguous. A researcher should be able to select trials at any of these levels or those involving particular categories of interventions, for analysis |
| B. Internal inconsistencies | ||||
| 2 | 2. Countries of Recruitment |
There were four fields (or pairs of fields) which provided clues as to whether a trial was Indian, Multinational, or Foreign: (1) Countries of Recruitment, (2) Recruitment Status of Trial (Global) and Recruitment Status of Trial (India), (3) Date of First Enrollment (Global) and Date of First Enrollment (India), and (4) Target SampleSize and Sample Size from India. Data in these fields was sometimes inconsistent Illustratively, even if the Country of Recruitment was only India, there were sometimes inconsistent values in the other 3 pairs of fields, such as non-zero values in the two Global fields |
Of these four fields (or pairs of fields) ClinicalTrials.gov only had Country (of recruitment). Therefore, for a global trial, there was no way to check the status of a trial in the USA versus in other countries |
Logic rules must be implemented to prevent contradictory information being entered in the database. Example 1: If India is the only Country of Recruitment, then Recruitment Status of Trial (Global) and Date of First Enrollment (Global) must become inactive fields, and Target Sample Size and Sample Size from India must have the same values Example 2: If India is not a Country of Recruitment, then Recruitment Status of Trial (India), Date of First Enrollment (India), and Sample Size from India must become inactive fields |
| 3 | 3. Relationship of Type of Trial and Phase of Trial | For the field Type of Trial, there were four options: observational, interventional, PMS, and BA/BE. Although we only chose interventional studies, the Indian set had 55 (3%) cases that listed Phase as PMS | Study type had 3 options: Interventional studies, Observational studies (including Patient Registries), and Expanded Access studies. PMS was not an option, and therefore we could not compare this field in the two databases | Logic rules must be implemented so that if Interventional is chosen for Type of Trial, then the PMS option of Phase of Trial should become inactive, thereby preventing a Phase from being chosen |
| 4 | 4. Confusion between PMS and Phase 4 trials | Trial registrants were sometimes confused about the terms PMS and Phase 4 trial. In these cases, PMS was chosen as the Type of Trial but the Phase of Trial as Phase 4 | Study type had 3 options: Interventional studies, Observational studies (including Patient Registries), and Expanded Access studies. PMS was not an option, and therefore we could not compare this field in the two databases | Logic rules must be implemented so that if PMS is chosen for Type of Trial, then Phase of Trial should become inactive, thereby preventing a Phase from being chosen |
| 5 | 5. Type of Trial: BA/BE versus Phase 1, 2, or 3 | Although in Type of Trial, BA/BE was a separate category, such trials were sometimes classified as Phase 1, Phase 2, or Phase 3 trials | There were 3 options in Study type: Interventional, Observational (subsection: Patient registries), and Expanded access. BA/BE was not an option, and therefore we could not compare this field in the two databases | Logic rules must be implemented so that if BA/BE is chosen for Type of Trial, then Phase of Trial should become inactive, thereby preventing a Phase from being chosen |
| C. Incomplete or non-standard information | ||||
| 6 | 6. Sites of study: incorrect listing of cities | In investigating the cities in which trials took place, we found incomplete or non- standard information in some cases. It appears that this information was entered in a free text field | It is not known whether the cities were listed correctly | Instead of being entered in free text fields, cities should be selected from a drop-down menu to ensure standardization of information |
| D. Missing data | ||||
| 7 | 7. Missing data |
A few other fields were also found to have missing data: (1) Name of PI was nil (2) Trials in Phases 1, 1/2, 2, 2/3, 3, 3/4, or 4 that had no information for Type of Study (3) Name of Primary Sponsor was nil (4) The state hosting the trial was not listed |
In earlier work [17], we found that PIs’ names were missing in many records, since it is a non-compulsory field. We have not done a systematic study of missing data in various fields | Most fields should be made compulsory, and unless data is entered in a particular format, sometimes from a drop-down menu, it should not be possible to register a trial. Ideally, the existing records should be updated |
| E. Variations in names or classification | ||||
| 8 | 8. Variations in a PI’s name |
There were variations in a given PI’s name, which sometimes made it difficult to unambiguously determine whether two names referred to the same person. Furthermore, two individuals may have shared a name |
In earlier work [17], we identified 19 categories of variations in PIs’ names. Some examples of the types of variations are listed belowc | Each PI’s name should be pre-registered, and thereafter, while registering the trial, the PI’s name should be chosen from a list through a drop-down menu. Since a person’s name may change over time, there should be a possibility to list the current name as well. Further, there should be a permanent and unique ID, such as an Open Researcher and Contributor ID (ORCID) number, linked to the name, with the system only accepting valid numbers |
| 9 | 9. The name and classification of the Primary Sponsor |
There were variations in a given organization’s name and classification Name of sponsor: For example, one company had the following name variants: BoehringerIngelheim India Pvt Ltd Boehringer Ingelheim India PvtLtd Boehringer Ingelheim India Private Limited Classification of sponsor: For example, one company may have been classified as “Pharmaceutical industry-Global” and as “Pharmaceutical industry-Indian” in different trials |
Name of sponsor: The Sponsor name seemed to be chosen through a drop-down menu, since each organization appeared to be represented by just one version of a name. By way of examples, each of the following organizations were listed multiple times in exactly the same format: Acotec Scientific Co., Ltd; Merck Sharp & Dohme Corp.; National Institute of Allergy and Infectious Diseases (NIAID); Albert Einstein College of Medicine, Inc.; Bausch & Lomb Incorporated Classification of sponsor: There was no field for sponsor classification. The closest field was “Funder type”, for which there were four categories: (1) NIH, (2) Other US Federal agency, (3) Industry, and (4) All others (individuals, universities, organizations) |
An organization’s name and its classification should be pre-registered, and thereafter, while registering the trial, the organization’s name should be chosen from a list through a drop-down menu Here, too, registrants should classify each trial using a multi-layered menu of options, along the following lines: The first layer should be a choice between Indian and foreign. The next should be between for-profit, not-for-profit, and government. The third layer should distinguish industry, hospital, research institution, university, charity, individual, and other, with these options being suitably available to the for-profit and not-for-profit categories, as applicable |
| F. Incomplete or incorrect details of Ethics Committees | ||||
| 10 | 10. Details of Ethics Committees | Every ethics committee did not have an address or clear institutional affiliation and may have been listed only by its name. The identity of such committees could not always be established unambiguously | There was no field regarding ethics committee approval, and therefore we could not compare this field in the two databases | The following details of each ethics committee connected to a trial must be spelled out clearly, regardless of whether it is an institutional or an independent committee: Name, affiliation (if applicable), and address. Each of these subfields must be pre-registered, and subsequently chosen from a drop-down menu while registering the trial |
| 11 | 10. Details of Ethics Committees | It was not always clear which site sought approval from which ethics committee | There was no field regarding ethics committee approval, and therefore we could not compare this field in the two databases | It must be possible to unambiguously identify which site sought approval from which ethics committee. Possibly the table listing the number of sites and the table listing the ethics committees could be merged |
| 12 | 10. Details of Ethics Committees | There were examples of trials for which there were significantly more ethics committees than sites, such as 28 committees for 7 sites | There was no field regarding ethics committee approval, and therefore we could not compare this field in the two databases | If the table listing the number of sites and the table listing the ethics committees were merged, it would prevent such irrational entries |
| 13 | 10. Details of Ethics Committees | There were examples of trial records in which foreign committees were included in the list of ethics committees, such as one in which 8 out of the 13 committees listed were foreign | There was no field regarding ethics committee approval, and therefore we could not compare this field in the two databases | Although the current guidelines for completing this field imply that only local ethics committees should be listed, this should be explicitly stated. Perhaps an information box should also reiterate this point while the trial is being registered. Also, if the table listing the number of sites and the table listing the ethics committees were merged, it would address this problem |
aAyurveda is a form of alternative medicine. AYUSH is an acronym for Ayurveda, Yoga, Unani, Siddha, and Homeopathy
b11 categories of interventions in ClinicalTrials.gov: Behavioral, Biological, Combination Product, Device, Diagnostic Test, Dietary Supplement, Drug, Genetic, Other, Procedure, and Radiation
c (i)Extraneous information with the name, such as prefixes, suffixes, or punctuation marks
(ii) Variations in the name in the form of spelling mistakes, different ordering of parts of the name, abbreviations of parts of the name, parts of the name missing, etc.
(iii) Other variations such as a name represented by just one word, or two people sharing a name