This prospective study was conducted at VRDL Government Theni Medical College, Theni, Tamil Nadu, India for six months from August 2018 to January 2019. The LRFs (n=645) evaluated in this study included all the investigative modalities offered in this VRDL. The pictograms for this study were designed by the author with the following characteristics; use of blue colour as in default LRF, clear and simple line art images, and use of plain large font both in English and Tamil (local) language. Each symptom in the clinical section was depicted by appropriate pictogram. This was attached onto default text oriented form as an overleaf. Therefore, for each patient, two types of clinical sections i.e., TLRF and PLRF were assessed. LRFs of both inpatient and outpatient received for routine laboratory investigations were evaluated. The permission for conducting this study was obtained from the Institutional Ethical Committee (IEC. No:884/MEIII/19).

The number of forms without clinical indication (blank), the different clinical information per form was calculated and the type of investigation in each form was also included. The duration of illness was depicted in pictographic scale that ranged from 1 to 10. In addition to that, the forms with duration of individual symptoms filled by the physicians were assessed. For the first of its kind, PAI was elucidated and calculated by the author which is the ratio of LRFs with the duration of symptoms that were filled by the physicians to the total number of LRFs received during the study period. The total number of symptoms per form and their type of symptoms were entered into a Microsoft excel sheet.

The data were summarised using a frequency distribution table and analysed using descriptive statistics. The response rate differences for TLRF and PLRF was depicted by proportional Venn diagram. All the data were expressed in percentages followed by its cumulative frequencies. Clinical information per patient entered in both the forms was considered as dependent variable. The difference in frequencies of categorical responses between two dependent groups was calculated by way of the comparison of proportions and chi-square. A two-sided p-value <0.05 was utilised as the criterion for rejecting the null hypothesis of no difference. All statistical analysis was done using SPSS software.

A total of 645 LRFs were evaluated, of which majority of the test done for dengue (622), followed by Hepatitis A Virus (HAV) (12), Chikungunya (8) and Rota virus (3). The number of LRFs with relevant clinical information increased significantly (p=0.0001) after the implementation of the pictogram, surpassing 85% compared to that of text oriented symptoms. [Table/Fig-1] shows the proportion of incomplete TLRFs received during the six months period. A small proportion of 2.2% (14) were left blank in both type of LRFs. Conversely, the duration of illness was filled more in TLRF than PLRF while 15.5% (100) were found unfilled in both forms [Table/Fig-2]. In the context of associated symptoms, more than one symptom per form was significantly (p=0.0001) improved in PLRFs. Majority of TLRFs (395) were with single or dual symptom(s)/form, on the other hand, in PLRFs frequencies under various symptoms showed almost a normal distribution [Table/Fig-3]. The maximum number of symptoms per request form was six and nine in TLRF and PLRF respectively.

[Table/Fig-1]:

Proportionate decrease in blank LRFs after implementation of pictograms.

[Table/Fig-2]:

Proportion of unattended duration of illness seen in PLRFs.

[Table/Fig-3]:

Comparison of total number of symptoms filled per LRF.

The proportion of each categorical response is tested by Test for Proportion (Fisher Exact test) and denoted as p-values 0.0001 (*), and 0.05 (#)

In addition, individual symptom wise frequencies were found to be higher in PLRF than TLRF; 4 vs. 7 (rash), 435 vs. 509 (fever) as shown in the [Table/Fig-4]. Significant improvement level at p=0.0001 was observed in knee pain-91.0% (101), rhinorrhea-88.6% (132), headache-81.2% (155). PLRFs attended by physicians mentioning the duration of symptoms was evidenced by [Table/Fig-5a]. Interestingly, PAI was 0.125 (21.7%) in PLRFs and this could not be possible in the default TLRF [Table/Fig-5b].

[Table/Fig-5]:

LRFs indicating Physician’s Attention Index (PAI).

Pictograms attracted form attendees and provide space for mentioning the duration of each symptoms (a) whereas the default TLRF (b) is designed with check box

Appropriate diagnosis and treatment provided to patients is always associated with symptom based clinical investigations as these symptoms will offer critical clues to the diagnosis. Even though the importance of appropriate completion of LRFs is usually emphasised for pre-registration house officers, the clinical history part of TLRF is often incomplete or even sometimes left blank [5]. The incompleteness in many of the TLRFs was in accordance with the study conducted at Lagos in which 1.3% of the 7,841 LRFs were fully completed [6]. In another case, TLRFs even fails to provide limited details of blood donors at Kano, Nigeria [7]. Salient works done by Salinas M et al., and Dogether MA et al., compared the completeness of paper based LRFs with LIS [2,8]. They found that 20-25% of the TLRFs were with blank clinical information. After implementation of LIS software based online Analytical Processing unit, the number of LRFs without clinical information were minimised. Similarly, this present study showed significant increase in completion rate of LRFs while replacing text oriented symptoms with pictorials.

Associated symptoms like headache, diarrhoea, abdominal pain is significantly filled in PLRFs. The pictorial form of symptoms not only attracts form attendees but also individuals without medical knowledge, feel ease to fill the form. This will be useful in most of the high patient load setting wherein physicians are unable to concentrate on each patient. Since this study was single blinded, the physicians were not advised or instructed to mention the duration of each symptom. So, the PAI in PLRFs is the direct indication of how pictograms enhance the involvement of physicians in form filling procedure.

The symptoms in default TLRF is grouped into syndrome such as diarrheal, respiratory, encephalitis etc. This kind of layout, if filled properly, will provide syndrome wise symptoms that aid the clinicians to narrow down the diagnostic tests and treatment protocol. In addition to that, majority of the TLRFs are filled with duration of illness and that again a crucial factor for diagnosing diseases like dengue. It is agreed that the conventional TLRF has stand out in some of the parameters but it under performs in clinical information section. In TLRF, repeated symptoms like fever, diarrhoea is mentioned under most of the syndromes which may lead to erroneous outcome. Hence, these symptoms may be merged into one symptom per clinical sign in PLRF. It is also noteworthy that the default TLRF doesn’t have provision to mention the duration in symptoms as of PLRF.

The utility of pictograms are mostly aimed to educate the patients on the use of medications and awareness programs in the form of leaflets or pamphlets namely in diabetes [9], haemodialysis [10], reflux Symptoms [11], psoriasis [12], dyspepsia [13]. The results of these studies showed that the patient information leaflets having pictorials of clinical signs have worked greatly in understanding the risk factors, transmission of disease and effective management of diseases. This is because, compared with abstract words, pictograms can make it easier for people to interpret and perceive information in a proper manner. This present study also brings out the evidence that imply pictorial symptoms in LRFs can greatly enhance the physician’s attention thereby improve the completeness of the request forms.

This study had certain limitations; firstly, not all symptoms such as jaundice, dark urine are converted into pictograms. Secondly, those clinical information that are filled in the PLRF are not verified with the patient’s case history. However, the outcome of this study suggests that the pictorial representation of symptoms in LRFs seems to be better and furthermore, the other sections has been taken into consideration for improvement in near future.

Combination of picture with text improves the quality in patient’s clinical history. Completeness in clinical information not only gives the laboratory personnel an opportunity to enhance patient benefit through post-analytical step interpretation but it also aid in the future retrospective studies. To put it in a nutshell, the authors emphasised the need to revisit LRF and to validate the usage of pictograms or any other novel yet attractive methods that would enhance accuracy and compliance with the necessities of laboratory request form. This, in turn, will improve the quality of patient/specimen information which aids the physician or diagnostic service providers to offer differential diagnosis.