Primary care serves as the bedrock of healthcare systems globally, acting as the first point of contact for individuals seeking medical attention. Understanding the most common diagnoses seen in primary care is crucial for healthcare professionals, policymakers, and researchers alike. Analyzing the reasons patients visit primary care facilities provides valuable insights into prevalent health concerns, healthcare resource allocation, and the evolving landscape of patient needs. This article delves into a robust study selection process designed to identify and rank the reasons for visits (RFVs) encountered in primary care settings, offering a comprehensive overview of the most frequent diagnoses.
To effectively determine the spectrum of common diagnoses, a meticulous study selection process was implemented. This process involved rigorous screening of titles and abstracts, followed by in-depth full-text reviews conducted by a team of three independent reviewers. The criteria for inclusion were specifically tailored to ensure the relevance and reliability of the data. Studies considered for review had to be set within general practice or primary care environments. A significant requirement was that each study reported a minimum of ten reasons for visits, ensuring a substantial dataset for analysis. Furthermore, to guarantee a representative sample size, each study’s population needed to encompass at least 20,000 visits or include data from a minimum of five clinicians over a year or more, or alternatively, track 7,500 patients over a year or more. This threshold was logically based on the typical patient volume seen in a standard primary care practice, approximating the patient encounters of five clinicians each seeing a reasonable number of patients daily. Observational study designs were prioritized to capture real-world primary care scenarios.
Conversely, specific criteria were established to exclude studies that might skew the results or not accurately reflect general primary care presentations. Studies focusing on very specific types of visits, such as routine periodic health examinations, were excluded to maintain focus on illness and symptom-driven visits. Similarly, research centered solely on acute conditions or specific health problems was not included, aiming for a broader view of primary care diagnoses. Studies that selected specific patient populations, like adolescents exclusively, were also excluded to ensure generalizability. A key exclusion criterion was studies where visits originated from referrals, such as to pediatrics or internal medicine specialists, as these do not represent initial primary care encounters. Finally, to ensure data relevance to contemporary healthcare practices, publications prior to 1996 were excluded. In instances where multiple publications drew data from the same source or database, the most recent and most comprehensive datasets were given preference. Duplicate publications were only considered if they offered unique analyses, such as subgroup analyses, enhancing the breadth of the review. Any disagreements during the selection process were resolved through team consensus or a third-party review to maintain objectivity. Efforts were also made to contact study authors to obtain any unpublished data that could enrich the analysis. For studies published in languages other than English, translation tools were utilized to ensure comprehensive review.
Data extraction was performed independently by two reviewers to minimize bias and ensure accuracy. The primary focus of data extraction was the reported reasons for visits (RFVs). These RFVs were defined as the specific reasons patients presented to primary care or the health problems managed by primary care physicians during these encounters. For each of the top RFVs identified in a study (up to twenty per study), detailed data was recorded, including the number, percentage, or rate of visits associated with each condition. Beyond RFV-specific data, descriptive characteristics of each study were also collected to provide context. This included whether the RFV was reported from the patient’s or clinician’s perspective, the total number of visits analyzed, the number of clinicians or practices involved in the study, the geographical location and duration of data collection, the proportion of female patients, the percentage of patients aged 65 years and older, and the coding system utilized for diagnoses (e.g., International Classification of Primary Care, ICD-9, ICD-10).
To evaluate the potential for bias across the included studies, a risk of bias assessment was conducted. Five key characteristics of each study were scored on a binary scale, with a score of 0 indicating a high risk of bias and 1 indicating a low risk. These characteristics encompassed the representativeness of the clinician sample (assessed by factors like gender diversity, practice experience range, and practice size variation) and the representativeness of the patient sample (considering gender balance, urban/rural mix, and age range inclusivity). Additional factors included whether data collection was prospective (low risk) or retrospective (high risk), whether a specific coding system was clearly specified, and the duration of data collection (one year or more indicating lower risk of bias).
The reported reasons for visits were categorized into “general” and “specific” categories to facilitate analysis. General categories represented broader descriptive groupings, such as “respiratory” issues. In contrast, specific categories referred to more precise diagnoses, like “pneumonia”. Within both general and specific categories, a standardized coding scheme was systematically applied to ensure consistency across different terminologies used in the original studies. For instance, within the specific RFV category, various terms like “back complaint,” “dorsopathies,” “back symptoms,” “dorsalgia,” “low back symptoms,” and “neck pain” were uniformly coded under the umbrella term “back pain/spinal pain.” Detailed diagnostic coding legends for both general and specific conditions were developed to maintain analytical rigor.
To analyze and synthesize the most common visit reasons, RFVs from each study were ranked by their frequency, from most to least common. Recognizing that studies reported visit frequencies in various formats (e.g., number of visits, percentage of visits, visit rates), the rank of each reported RFV was chosen as the standardized measure of relative frequency. Using the top 20 ranked RFVs from each study, a scoring system was applied: the most common condition in each study received a score of 20, the second most common a score of 19, and so on, down to a score of 1 for the 20th most common. Any RFV not within the top 20 in a given study received a rank of zero for that study. These rankings were then combined across all studies, and mean ranks were calculated for each RFV. The RFVs with the highest mean ranks were identified as the most commonly seen reasons for visits in primary care. To ensure robust results, any RFV that appeared in only a single study was excluded from the combined analysis.
In addition to identifying the most common diagnoses, secondary analyses were conducted to explore other dimensions. Countries included in the studies were categorized by their economic status as either developed or developing, based on the United Nations classification system. This categorization allowed for a comparison of the mean ranks of clinician-reported RFVs between developed and developing economic regions, exploring potential differences in primary care visit patterns based on economic context. Furthermore, when subgroup analyses were available from the included studies (e.g., based on clinician or patient sex, or practice setting), data from these subgroups were combined using the same mean rank approach, provided that at least two studies offered data for a specific subgroup. This comprehensive analytical approach allowed for a nuanced understanding of the most common diagnoses seen in primary care and the factors influencing these patterns.
