Conducting Record Review Studies in Clinical Practice

Clinical record review or chart review is a previously recorded data to answer clinical queries. Such a study can be used to answer specific clinical questions in a relatively easy and less resource intensive manner. But these studies may be constrained by the limited information retrievable and inadequacy of records. Various types of data sources may be available for conducting such reviews (like case charts, computerized registries, etc), each with specific strengths and weaknesses. The procedure usually consists of drawing up the research question, identifying the appropriate data source, devising a data extraction plan, extracting the data, checking for errors, data analysis, and appropriate archiving and dissemination of the findings. The ethical aspects in such studies primarily pertain to issues of informed consent and confidentiality. This paper provides a broad overview of how to go about a clinical record review, and serves as a ready reference for those who would like to undertake such record reviews.

Clinical record review is a process aimed at obtaining retrospective data to answer clinical queries. It is also known by other names like ‘retrospective data analysis’, ‘clinical chart review’, ‘chart review’ and so on. It is a useful method when data has been recorded in case notes or a structured database and an analysis of the data elements needs to be done. It entails studying data that has already been recorded and involves summarizing the data, subjecting it to appropriate statistical analysis and drawing inferences. It has been widely used in various clinical disciplines like medicine [1], pediatrics [2], orthopedics [3], psychiatry [4],and dermatology [5], to name a few. Record reviews have extended and consolidated the scientific evidence base by evaluating disease characteristics and course over time as well as treatment outcomes [6,7]. The other reasons for carrying out a clinical record review could include determining the appropriateness of diagnoses, problem identification, treatment and care planning, and to assess adherence to guidelines’ standards [8,9].

There are specific differences and challenges in performing clinical record reviews as compared to prospective studies. Prospective studies are able to plan what data to gather and how to obtain the information. Gathering selected data may be sufficient for prospective studies where primary data gathering is based upon a hypothesis. Record reviews, on the other hand, are based upon information that is already available in the form of records. If particular information is not recorded reliably, then it cannot be used for record reviews. Hence, prospective studies can have a more focused approach compared to record reviews. Also, data gathering can be more systematically standardized in prospective studies, than in record reviews. Broadly, prospective studies seem to be more appropriate to demonstrate causal relationship, while record reviews may be more helpful in finding associations.

Conducting a record review has several advantages. This kind of study requires less effort and time compared to prospective studies. It is therefore less resource intensive. It enables assessment of a large sample at limited cost. It enables easy collection of information which is routinely recorded. It minimizes recall bias for an event in the past. It also reduces the need for intrusion into patients’ time for assessment as part of the study. For some types of data, record review may be the most feasible type of study. For example, knowing the prescription trends of a particular medication in a country [10]. On the other hand, record reviews have certain disadvantages. Variation in the manner in which data has been gathered and recorded in the charts limits the extraction and interpretation of the variables. Some records may be incomplete or lost in the course of time, leading to missing data. Also, records may not have been stored in an easily retrievable manner restricting the extent to which they could be utilized further.

This paper deals with record reviews conducted in clinical practice and delves into the various data sources that can be utilized, the methodology for conducting a review, the utility and applications of such reviews and ethical aspects pertaining to these reviews.

A variety of data sources can be utilized to collect information for a record review [Table/Fig-1]. These sources could be in the form of hard copies of case notes and case files, manually entered registers, and computerized databases [11–14]. Data sources can include case notes, inpatient case files, attendance registers, nursing records, pharmacy records, disease registries (like cancer registries), laboratory records, adverse event monitoring systems, clinical trial information, national demographic records and so on.

[Table/Fig-1]:

Data sources for clinical record review

Each type of data source presents unique opportunities and challenges. Inpatient case files may provide a detailed account of the patient’s symptoms and evolution of the disease, but may be exhaustive and require time consuming data extraction by trained staff. Disease registries on the other hand provide robust data about the demographic and clinical presentation of diseases. However, these data sources require good reporting services to be a valid representation of the disease in question. Outpatient attendance registers and pharmacy records may contain certain details of a large sample of a variety of patients, but they usually provide information about a very limited number of variables. The national demographic databases (for events such as suicide, etc) are reservoirs of information about a large population, and are representative of the community sample. But these extensive repositories require permissions and clearances before use for academic purposes. Also, sifting through such large databases may require extensive and dedicated computing time.

The type of data source would determine the extent and type of information that can be gleaned from it. The selection of the data source is based upon both convenience as well as the requirements of the clinical query. Certain extensive data sources may require significant resources and large organizational supports for processing. On the other hand smaller databases may provide information about a limited sample. Also, some data may be missing or become non-retrievable with any type of source, depending on the manner of recording and storage of data. Hence pragmatic and practical considerations quite often determine the usefulness of a data source for purposes of answering a clinical query.

The process of clinical record review involves multiple steps [Table/Fig-2]. The exact procedure and protocol needs to be modified in accordance with the type of study, nature of the data source and local constraints [15–18]. Some of the basic and commonly followed steps are as discussed as under: Formulating the clinical question. The first and foremost step is to clearly formulate the clinical question which needs to be answered through the record review. For example, what are the characteristics of patients attending a psychosexual clinic, which medication regimen is associated with better adherence to treatment among adult diabetes patients and so on. The clinical question can broadly follow the ‘PICO’ model, wherein it is useful to state beforehand the patient population being studied; the intervention, investigation or the characteristics of the sample under study; the control population or condition if any; and the outcome of interest. The more precise the clinical question, the less vague are the results. It must be noted that preliminary and descriptive studies might use a less defined population or control condition to answer their clinical query.

Quality assessment of record reviews is another important research consideration. Presently there are no uniform guidelines for reporting of chart reviews as compared to the ones developed for reporting meta-analysis (PRISMA) and randomized controlled trials (CONSORT) [19,20]. Yet, various methods have been suggested for the assessment of quality of record reviews. These have included checklists to assess the various methodological aspects of record reviews [21,22]. Briefly, checklists contain elements like representativeness of the sample, tackling inconsistent data, avoiding misclassification bias, declaration of conflict of interest of the authors and so on. Abstraction methods and tools have been devised for accurately retrieving data from chart reviews [21,23]. It has been suggested that vague definitions, imprecisely worded research questions, and poor initial data collection can impair the quality of the information gathered from the record reviews [23].

Record reviews have found widespread application in the clinical setting and include:

The application of record reviews for clinical purposes may be varied. The above are some representative scenarios. Factors which contribute to a successful and relevant record review include a genuine clinical need, good availability of records and research enthusiasm.

Informed consent and patient confidentiality are the important ethical issues relating to record reviews [23,30,31]. Usually when data is collected and stored in medical records, it is not done so with the explicit intention of further use in research in most situations. Hence, there is no a-priori informed consent for the use of data in a subsequent record review study. Using such data without the consent of patients for research may be considered ethically questionable by some authorities. There is the possibility that confidential data of the patient may be dealt with and possibly misused, which may jeopardize the doctor-patient relationship. Sensitive data about the patient may fall into hands of those not directly involved in patient care or those not bound by the requirement of patient-doctor confidentiality. For example, data about HIV status of a patient may be accessed in a study of gynecological surgery outcomes, the unwanted disclosure of which may lead to difficulties and distress for the patient. Hence, caution should be exercised while using information from records.

In the scenario of record reviews, following certain principles can help in ethical conduct of the study. Firstly, only that much information should be extracted and coded as is required for answering the research question. Usually, informed consent is not required beforehand; if no more than routine clinical information is used for the analysis. Secondly, any identifying information from the data set should be removed by data controllers before further usage and analysis. The data (both variables and their values) may be coded into alpha-numeric format for concealment with few designated persons having the coding key. For example, the variable of high blood pressure can be coded as ABC001, with values of 0 or 1 representing having hypertension or not. Thirdly, safeguards must be in place for appropriate and ethical use of the data. Confidentiality clauses may be explicitly specified for those who do the data extraction. It should be ensured that data, especially in electronic format, are accessible to authorized personnel only, and appropriately archived or deleted. Fourthly, due consideration should be given towards seeking ethical clearance from an Institutional Review Board (IRB) before starting data collection, especially while dealing with sensitive information. Many IRBs have policies of exempting certain types of record reviews, but this should be clarified prior to conduct of the study.

Whether to trace the subject/ patient to disseminate the clinical findings poses another ethical question (especially those which are potentially beneficial for the subject/ patient in question). It may be prudent to weigh the risks and benefits to the patient in such situations. But in most situations it would be appropriate to restrain the enthusiastic tracing of patients by research investigators, especially when formal informed consent has not been obtained.

There are certain potential areas where errors may crop up in record reviews. The clinicians and researchers may inadvertently miss out important components while reporting record reviews. Some of these are summarized in [Table/Fig-3]. It may be acknowledged that the description of methodology while presenting the final findings of the review is often constrained by the limited space available for the manuscript in the journal of publication. Yet, a clear description of the methodology encompassing the crucial points should be presented in a coherent manner.

For addressing the shortcomings of chart reviews, there is a need for standardized objective assessment measures for their reporting. The reporting of randomized controlled trials and systematic reviews has benefitted from the use of guidelines and brief checklists [32]. It can thus be surmised that reporting of chart reviews may benefit from the development of standardized reporting guidelines in the future.

Record reviews are useful tools for gathering data about medical and surgical illnesses, investigative procedures and their outcomes. The effective utilization of record reviews for seeking answers to specific clinical questions requires adequate planning and use of appropriate data sources. Record reviews offer several advantages in their conduct but their results should be interpreted keeping in view their limitations. Ethical issues, especially those pertaining to confidentiality of records, need to be taken into consideration. Appropriately conducted record reviews can help in effectively expanding the scientific knowledge base, and can provide information which may not be available through other means.

[1]. Page KR, Sifakis F, Montes de Oca R, Cronin WA, Doherty MC, Federline L, Improved adherence and less toxicity with rifampin vs isoniazid for treatment of latent tuberculosis: a retrospective study Arch Intern Med 2006 166(17):1863-70.  [Google Scholar]

[2]. Curtin C, Bandini LG, Perrin EC, Tybor DJ, Must A, Prevalence of overweight in children and adolescents with attention deficit hyperactivity disorder and autism spectrum disorders: a chart review BMC Pediatr 2005 5:48  [Google Scholar]

[3]. Fernandez MC, Gottlieb M, Menitove JE, Blood transfusion and postoperative infection in orthopedic patients Transfusion 1992 32(4):318-22.  [Google Scholar]

[4]. Basu D, Sarkar S, Mattoo SK, Psychiatric Comorbidity in Patients With Substance Use Disorders Attending an Addiction Treatment Center in India Over 11 Years: Case for a Specialized “Dual Diagnosis Clinic.” J Dual Diagn 2013 9(1):23-9.  [Google Scholar]

[5]. Kaufmann MD, Desai S, Special requirements for electronic health records in dermatology Semin Cutan Med Surg 2012 31(3):160-62.  [Google Scholar]

[6]. Curtis JR, Yarnold PR, Schwartz DN, Weinstein RA, Bennett CL, Improvements in outcomes of acute respiratory failure for patients with human immunodeficiency virus-related Pneumocystis carinii pneumonia Am J Respir Crit Care Med 2000 162(2 Pt 1):393-98.  [Google Scholar]

[7]. Shahedi K, Fuller G, Bolus R, Cohen E, Vu M, Shah R, Long-term risk of acute diverticulitis among patients with incidental diverticulosis found during colonoscopy Clin Gastroenterol Hepatol 2013 11(12):1609-13.  [Google Scholar]

[8]. Cheng CL, Kao YH, Lin SJ, Lee CH, Lai ML, Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan Pharmacoepidemiol Drug Saf 2011 20(3):236-42.  [Google Scholar]

[9]. Aggarwal NK, Rosenheck RA, Woods SW, Sernyak MJ, Race and long-acting antipsychotic prescription at a community mental health center: a retrospective chart review J Clin Psychiatry 2012 73(4):513-17.  [Google Scholar]

[10]. Mertens JR, Lu YW, Parthasarathy S, Moore C, Weisner CM, Medical and psychiatric conditions of alcohol and drug treatment patients in an HMO: comparison with matched controls Arch Intern Med 2003 163(20):2511-17.  [Google Scholar]

[11]. Ehrenstein V, Antonsen S, Pedersen L, Existing data is for clinical epidemiology: Aarhus University Prescription Database Clin Epidemiol 2010 2:273-79.  [Google Scholar]

[12]. Lau EC, Mowat FS, Kelsh MA, Legg JC, Engel-Nitz NM, Watson HN, Use of electronic medical records (EMR) for oncology outcomes research: assessing the comparability of EMR information to patient registry and health claims data Clin Epidemiol 2011 3:259-72.  [Google Scholar]

[13]. Mertens JR, Weisner C, Ray GT, Fireman B, Walsh K, Hazardous drinkers and drug users in HMO primary care: prevalence, medical conditions, and costs Alcohol Clin Exp Res 2005 29:989-98.  [Google Scholar]

[14]. Jensen PB, Jensen LJ, Brunak S, Mining electronic health records: towards better research applications and clinical care Nat Rev Genet 2012 13:395-405.  [Google Scholar]

[15]. Zwaan L, de Bruijne M, Wagner C, Thijs A, Smits M, van der Wal G, Patient record review of the incidence, consequences, and causes of diagnostic adverse events Arch Intern Med 2010 170:1015-21.  [Google Scholar]

[16]. Cassidy LD, Marsh GM, Holleran MK, Ruhl LS, Methodology to improve data quality from chart review in the managed care setting Am J Manag Care 2002 8:787-93.  [Google Scholar]

[17]. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J, Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med 1996 27:305-08.  [Google Scholar]

[18]. Reisch LM, Fosse JS, Beverly K, Yu O, Barlow WE, Harris EL, Training, quality assurance, and assessment of medical record abstraction in a multisite study Am J Epidemiol 2003 157:546-51.  [Google Scholar]

[19]. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration BMJ 2009 339:b2700  [Google Scholar]

[20]. Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomized trials Ann Intern Med 2010 152(11):726-32.  [Google Scholar]

[21]. Liddy C, Wiens M, Hogg W, Methods to achieve high interrater reliability in data collection from primary care medical records Ann Fam Med 2011 9(1):57-62.  [Google Scholar]

[22]. Kaji AH, Schriger D, Green S, Looking Through the Retrospectoscope: Reducing Bias in Emergency Medicine Chart Review Studies Ann Emerg Med 2014 pii(S0196-0644(14)):00293-5.  [Google Scholar]

[23]. Allison JJ, Wall TC, Spettell CM, Calhoun J, Fargason CA Jr, Kobylinski RW, The art and science of chart review Jt Comm J Qual Improv 2000 26(3):115-36.  [Google Scholar]

[24]. Zhang T, Zhu Q, Zhang X, Ding Y, Steinhoff M, Black S, The clinical characteristics and direct medical cost of influenza in hospitalized children: a five-year retrospective study in suzhou, china PloS One 2012 7(9):e44391  [Google Scholar]

[25]. Larsson M, Nguyen LHT, Wertheim HF, Dao TT, Taylor W, Horby P, Clinical characteristics and outcome of Penicillium marneffei infection among HIV-infected patients in northern Vietnam AIDS Res Ther 2012 9(1):24  [Google Scholar]

[26]. Zaide GB, Pekmezaris R, Nouryan CN, Mir TP, Sison CP, Liberman T, Ethnicity, race, and advance directives in an inpatient palliative care consultation service Palliat Support Care 2013 11(1):5-11.  [Google Scholar]

[27]. Bates DW, Spell N, Cullen DJ, Burdick E, Laird N, Petersen LA, The costs of adverse drug events in hospitalized patients.Adverse Drug Events Prevention Study Group JAMA 1997 277(4):307-11.  [Google Scholar]

[28]. Lewis PF, Schmidt MA, Lu X, Erdman DD, Campbell M, Thomas A, A community-based outbreak of severe respiratory illness caused by human adenovirus serotype 14 J Infect Dis 2009 199(10):1427-34.  [Google Scholar]

[29]. Hayward RA, Hofer TP, Estimating hospital deaths due to medical errors: preventability is in the eye of the reviewer JAMA 2001 286(4):415-20.  [Google Scholar]

[30]. Haynes CL, Cook GA, Jones MA, Legal and ethical considerations in processing patient-identifiable data without patient consent: lessons learnt from developing a disease register J Med Ethics 2007 33:302-07.  [Google Scholar]

[31]. McCarthy RL, Ethics and patient privacy J Am Pharm Assoc 2008 48:e144-5232.  [Google Scholar]

[32]. Moher D, Jones A, Lepage L, CONSORT Group (Consolidated Standards for Reporting of Trials). Use of the CONSORT statement and quality of reports of randomized trials: a comparative before-and-after evaluation JAMA 2001 285(15):1992-95.  [Google Scholar]