DATABASE MANAGEMENT PI SURVEY RESULTS - PowerPoint Presentation

Slide1

DATABASE MANAGEMENT PI

SURVEY RESULTS

What are your biggest concerns about data management for your research protocols?

I do think data managers need better managers than just PIs--ideally masters or PhD level biostatisticians who can make sure they are doing things correctly.

Lack of interoperability our various data repositories. Administrative overhead inhibiting projects. Lack of interest in supporting research applications of data mining projects (compared to financial management and clinical support resources).

Integrity Longevity Completeness Security Access (no pun intended) to the data

So far, with mostly chart review projects, only simple databases have been necesary, for which no dedicated manager has been needed.

it is a very expensive undertaking. If there were a centralized source it might help to cut down on costs

Managers for hire do not have content expertise We need a uniform system integrating data collection into databasing

Consistency with turnover of data management staff

Stability over time Ease of access

Correct entry of data, correct interpretation of clinical events as it is entered into the database.

That i do not adequately utilize the managers, and that they might enter data without recognizing it is incorrect.

JHU HR

payscale

is very low compared to what these positions would be paid in the private sector, making it very difficult to compete for top quality candidates. Servers that house our data are managed by folks who do not fully appreciate the importance of data integrity and data security. Our data our housed on three different servers and they all have different policies for backing up data, some only saving data back to six weeks. More sophisticated IT support would be helpful.

The resources needed to support a data core on an ongoing basis are enormous and would take up a very large chunk of typical NIH/foundation grant budgets. This makes fundraising challenging for a data core.

maintaining

searchability by discrete elements of data vs. narratives that are not searchable; inability to "cross populate" data from other databases, necessitating transcription of data already available, which is both time consuming and error prone.

Data systems within JHU/JHH do not interface. Licenses for software held by

JHM,run

over the internet are not kept up to date, such as

SPSS.There

are days to weeks when databases are not available because of this issue.

Data entry errors and lack of knowledge of program

Ability to get data base management assistance for preliminary data and without funding for a person to assist. I want to use

RedCap

and don't have the resources to use that data base. If my grant is funded my collaborator has free access to it. I would like to see this database

recieve

IT support via the ICTR.

insufficient experience

Continued access to highly skilled personnel for high level data management/database modification and programming.

Audit trail robust relational database

There is no financial, human or computer resouces that I can access to adequately construct a fully functional 21st century clinical database within my control.

Finding people to design a database and how to pay them for periodic adjustments changes to the database.

Finding one willing to work on a project part-time or institutionally reported.

My studies are relatively small at present and I feel I can manage the data myself using SPSS. That said, I also have little choice in the matter given the unfunded nature of the work right now.

Lack of additional training opportunities.

I have not used databases for my research so far.

Access to data mangers and timeliness of reports.

Merging of data from multiple institutions

I don't even know the resources available for data management so I do it myself.

The new expectations that our databases need to be compatible with

CaBig, CaTissue etc.

That the data are entered accurately, and that the data base manager is double checking his/her work.

Identifying qualified data managers. We have a superb one now.

I would be lost if i ever lost the person I now have. As my answers to questions 7 and 8 show, I am very lucky to have the person I have (who was actually hired by my then-project director over 10 years ago). If i had to replace her, I'm not sure I would know how to begin. Not only does she do a fantastic job on our research, she teaches our post-docs about data management. She does all of this long-distance, as she has lived in western Pennsylvania for the past 5 years. Therefore, any new ICTR opportunities for networking/training would need to take this into account.

It would be great if I had sufficient funds to have one person dedicated to data management! I can't imagine a universe with this type of luxury.

Inability to monitor accuracy due to time constraints

Data security and integrity

I am very interested in meeting with other data managers. I have a strong technical background, but no experience in research, and would really appreciate a network of people with whom I could trade information and experience.

Currently doing it all myself. Concern for completeness of data acquisition given time constraints, etc.

Data entry errors, formatting of data entered so that data could be used for biostatistical analyses.

Availability of trained personnel. Hard to believe there's an employment crisis in the US. Plenty of data coordination, data QA, and regulatory-compliance positions to go around!

There are not enough database managers who are available contractually to help with development and running data for trials. This requires dedication and detail oriented people who are often hard to find. Department budgets don't usually have the money to support their own data manager.

insufficient interface with/access to GI tissue bank and database

Making sure that data is not lost when program encounters errors (File-maker PRO data base).

It was difficult to answer question 3 above, as we are part of two national consortia, with some data management here and some by the data coordinating center at Mass General Hospital. My biggest local concern is our general lack of easily accessible tools to quality check data entry. I am fortunate to have careful individuals working on our data collection.

Turnover of personnel who control. Proper backup of information. Procedures to minimize data entry errors.

Any potential violations of privacy

managers maintaining skills IT support lacking knowledge to adequatley supervise of data managers

Support for web-based surveys-- we had a hard time identifying who to call and then uploading it to the surver. We could use more support in this area. Ensuring that databases are built correctly to ensure that data is accurately captured and easy to clean and manipulate for analyses. This is particularly true for data directly imputed by patients. Would like to move into audio CASI, but don't have the staff or resources to do so. Important since we work with low-literacy participants.

We need to create the database from paper source documents. Dr. Diener-West will assist us in the selection and supervision of the candidate.

I have been careful about hiring competent data managers and working to retain them. The best are too easily hired by local pharmaceutical companies or CRO's, who can pay 2-3 times in salary and bonuses what we pay.

Quality of the data captured

continuity; position turnover can be an issue

I have no main concerns. A minor issue is making sure the data are entered in a format that can easily be read by a statistical software package.

I do not currently have a database as part of my research protocol

advance programming to extract and organize data for analysis

efficiency

It is a time-consuming role that is vastly under-estimated and under-funded by NIH. There is little training, at the JHU institutional level, regarding data management and quality assurance issues for investigators and research staff -we "invented the wheel" to create our own systems, but there may have been more efficient ways that we could have learned along the way.

1.

Security. 2. Loss of data 3. Lack of access to data

My only concern is that the current process will pull my data management staff away from current work, and have them attending something else that becomes a mandatory training, etc., by the university.

I am very happy with the employees who have been doing our database management. Honestly the biggest concern is trying to piece together funding to continue to support them while the grant funding is reduced. I have a great employee who has been working with our group for more than 10 years, and I may have to let him go soon if we can't find some additional funding for him.

I think it's been great so far. I'm a little worried about problems with my statistician not being able to access the database, but we're working on that.

We do not have established SOPs for data management and each study seems to be reinventing the wheel. Given the integration of data management with QA issues, this is critical issue as it impacts data quality.

My next grant attempt involves a much larger data field. At that time, I do anticipate using a manager to monitor my data. To date, I have not needed a database manager.

Being able to find good technical assistance. As I am not fulltime reserach nor NIH-funded, I rely on exisiting divisional resources or paying a la carte for data management/statistical support needed for my studies.

Funding

not a stanrade process..we worked with two different groups for our very small project before we decided to do it ourselves. first was solo contractor...she did not do the work in a timely fashion, second was a group from an external university in mid west (i dont know theier names) they also ran out of steam, became unresponsive to our needs for skip patterns that worked in teh questionaire we diesigned and we terminated that agreement. WE now use a Survey monkey product for data collection and excell and SPSSS to analyze data.

My own limited training in this area and need to rely on colleagues.

If I ever get funding for my own independent research, I do not know who will manage the database. It would be nice to have a pool of available database managers for protocols for whom salary support could be provided in grant applications which might help with obtaining funding.

The lack of centralized and

consistant

HIPAA compliant database development and management for investigator initiated trials

Intergration

of database management and statistical analysis.

1) Being able to support a sufficient level of effort to get the task done. 2) Being able to get biostatistics consultation for the database manager to maximize the efficiency of the system.

understanding social science research and data needs for project determining skill level

Lack of oversight

the learning curve the payscale and ability to move up some are lifers - others use job as stepping stone - so -turnover can be a concern

Lack of resources for quality control of data quality

database crashes missing data missing datafields difficulty with search items not forseen

We are fortunate to have an extremely competent database manager, so at this time, I have no concerns.

Our data volumes are enormous and it will pretty soon become difficult for single groups to keep them at a high level

Not enough time to do myself and not enough resources to hire anyone else.

Quality COntrol

access to the data routine QC of data entered

1. How to ensure the data are entered accurately, up to date, and clean/ready for analysis. 2. How to export data from Access and import it into SPSS. 3. Need for more dynamic data entry system (e.g., REDCap) and folks with expertise in that system to assist. 4. Need for an efficient data management system to do tasks such as combining distinct datasets that have common data elements.

It would be helpful to have educational sessions on how to assess database managers, supervise them, as well as support to improve their skills through educational opportunities. Thank you.

Not a problem for me: large-scale work done by coordinating centers; small-scale work done by junior faculty, fellows, and students who work with me, all of whom have a least masters level training in

epidmeiology, including some database experience.

Need more consistent quality control measures

Lack of training Lack of database development support from the ICTR

Mistake during data entry.

Accuracy of the data that is put into the database. Ability to employ someone that can put data into the system.

Understand the importance to complete the data entering and monitoring in a timely fashion - within a couple of days of data collection

quality of data collection and management at overseas sites

I have no concerns since data management for our research protocols are managed by our sponsors and they seem to do an excellent job managing our data.

Ability to anticipate future needs Ability of platform to last for years (eg will I be able to retreive this data 20 years from now?) Ability to store imaging studies

organization of data

Data organization and storage

My biggest concern is that someone takes the data from JHH email system.

Clinical knowledge of the speciality being studied. Reliability measures for the data. Flexibility within collection process for future data. Error rate.

I have none and am quite comfortable with the analysts at UWash.

Losing current staff and needing to replace them.

Developing

stratagies

that will ensure that we are collecting the most valuable data.

Data verification Privacy

Majority of data management I have done has been with fellows.

No supervision by data mgt mavens; no community; no opportunity to advance. Thus, best and brightest don't want to work here.

My lack of experience.

Managing diverse types of data

It has been difficult to identify a data manager with adequate data management skills since my study budget does not allow for a FT data manager. Also, there seems to be limited sharing of "best practices" for data management.

compliance to IRb protocol

- Difficulty in obtaining timely, accurate follow-up data - Multiplicity of systems at JHMI that don't talk to each other - Inability to readily and securely link to data from JHMI clinical databases, and to extract information from free text in medical records - Difficult to fund an individual just to be a database manager, so they tend to have other responsibilities that can distract them - Database managers need more opportunities for advancement and staying abreast of latest technologies

Multiple clinical systems for source documents make it difficult to appropriately and accurately gather data, especially in a short time window, as is often required by sponsors. Similarly, CRFs are not standardized, and I feel these varied reporting systems increase the chance of error and delay in data management

The lack of specialization and training in this important area. Essentially these people learn through trial and error, become highly competent at a single system, and show little ability (generally) to work across systems. The lack of a unified approach through the SOM does not help.

I do not conduct research at this point in time; the studies I do conduct are educational in nature and do not generate a database that requires a manager

Use of Access for data management without good tracking of modifications of the data

back up

Following protocols exactly as specified, collecting all necessary data, recruiting, etc

Center for Clinical Trials is competent but startup time is very long and this delay is very costly to the trials I am working on. Frankly I'd like to develop our own local competency and drop them.

None to mention, My projects are relatively small and I have some basic data management skills.

None

I don't have any concerns now that I have hired this person. That said, I am completely dependent upon his expertise.

Integrety of data; preventing inadvertent and untraceable changes; increasing knowledge of senior colleagues who think they understand database needs/requirements from imposing their ideas on staff.... data security writing adequate sections in grant applications about attending to database oversight

Lack of adequate funds for a data manager

No concerns.

data integrity

Accuracy Statistical analysis

none at this time

Time is the most critical element and it would be a major support to have someone assist with database management.

adquate medical knowledge

consistency, ease of use and robustness to meet individual needs

I have managed the data for my protocols. This might change with my new RO1 grant.

That they are performed accurately.

IRB and protocol compliance

Our biggest concern about data management is how to support/pay for this person?

Need help developing better databases which allow for research and meeting requirements for monitoring patients and reporting AE's.

That I haven't paid enough attention to data management. I am hoping to use RedCap in the future which should help with some structural issues

support

I am highly dependent on the managers with whom I work and if they were ever to leave JHU I would have difficulty replacing them. A centralized service would likely work better. However, there are smaller projects with limited funding (at least early on) that require some database management input and if the service were too expensive these might not be viable.