Middle East Journal of Family Medicine

BA Al-Mustafa, Qatif-3 PHC Center.
AR Ghulam, formerly of Vaccination Unit, Department of Epidemiology, Qatif Primary Health Care.
GM Al-Qatari, Director, Qatif Primary Health Care.
AA Al-Sinan, Department of Pediatrics, Qatif Central Hospital.
HM Al-Hani, Infectious Diseases Unit, Maternity and Children Hospital, Dammam.
AM Al-Omran, Primary care department, National Guard, Dammam.

Correspondence to:
Bader A. Almustafa, MBBS, DPHC (RCGP), ABFM, SBFM, ECHS
Consultant family physician,
Qatif-3 Primary Health care,
P.O. Box 545,
Qatif 31911, Saudi Arabia
Tel. +966 3 854 1436 ext. 160,
Fax. +966 3 852 6834,
E-mail: bader@alqatif.org

ABSTRACT

Background: Reporting of adverse events following immunization (AEFI) is very low. This might be related to low recognition and documentation by the practicing physicians.

Objectives: To estimate the level of documentation of vaccination history and its predictors for hospitalized children of = 2 years at the time of clerking.

Settings: Hospitalized children of = 2 yrs admitted in 4 governmental and private hospitals in Eastern Saudi Arabia.

Results: Out of 299 admission notes, 91.6% used were in pre-designed forms. Detailed vaccination history was documented in 1.7% only. "Up-to-date vaccination" statement was common in 58.5%, while 39.8% contained no vaccination history at all. Documentation tends to be more succinct in pre-designed admission forms holding a predefined field for vaccination history. Authors identified 18 (5.8%) cases with possible AEFI which were missed by the practicing physicians.

Conclusion: Vaccination history is poorly documented in admission notes, which might reflect poor recognition and reporting of AEFI by the practicing physicians. Improving the documentation may result in better reporting of AEFI.

Keywords: adverse events; vaccination; Saudi; chart documentation.

A wide range of adverse events follows vaccination. These include different local, behavioral and systemic effects, which vary from frequently occurring minor effects to occasional serious effects¹. Monitoring of adverse events following immunization is essential and beneficial as shown in Box 1².

WHO has recommended that all cases requiring hospitalization that occurred within a month of immunization to be reported. 3 However, such monitoring relies on spontaneous reporting by clinicians, which is very low in many countries, including Saudi Arabia, where it, seldom, approaches zero^4,5. This might be related to low recognition of the problem, as a possible indication for hospitalization^6,7. This low recognition might be reflected by the level of documentation of vaccination history at clinical encounters. Our search failed to trace any work done to assess the level of documentation of vaccination history in pediatric admission notes. However, Philip Kum-Nji and colleagues in Tennessee showed that admitting physicians failed to document the immunization status of 22% of the pre-school patients⁸.

Objectives
This study aims to estimate the level of documentation of vaccination history and its predictors for hospitalized children of = 2 years at the time of clerking.

Two private and two governmental hospitals were selected from 13 private and Governmental hospitals working in the main cities of the eastern province of Saudi Arabia. Ten percent of pediatric admittances to the medical pediatric wards were selected using systematic random sampling.

Children transferred from other pediatric wards such as nursery, intensive care units, and surgical wards were not included. Medical records of children of 2 years and below admitted in the period March 2001 - March 2002 (Hijri year 1422) were reviewed. Data were collected from clerking sheets and discharge summaries of every admittee, by the authors. Data collected included demographic data, level of documentation of vaccination history, discharge diagnoses, whether a pre-printed form has been used for clerking and whether a specific field for vaccination is pre-defined.

Documentation of last vaccination shot was categorized as either "detailed" when type and timing of the shot was documented in clerking sheet; "up-to-date" when statement of up-to-date was documented only; or "no vaccination history" when no history related to vaccination was documented. Three of the authors, individually, reviewed discharge diagnoses and categorized them upon their likelihood of being vaccine-related, as defined by WHO^2,3. Diagnosis was categorized as likely, unlikely or indeterminate once agreed by two authors out of the three.

Categorical data were cross-tabulated, while continuous data were recorded into groups of interval.

Data were tested for significance, using Pearson Chi Square²? and Fisher's exact test, where applicable. Stepwise multinomial logistic regression was used to predict the level of documentation of vaccination history, while binary logistic regression was used to predict the documentation of detailed vaccination history (DVH) and to predict no documentation of vaccination history as well. Intra-hospital data analysis was carried out after controlling for name of the hospital. A p-value of < 0.05 was considered statistically significant. Statistical Package for Social Sciences version 11 was used for analysis.

Out of 303 medical records reviewed, four were excluded due to incomplete data. The characteristics of the 299 medical records included are shown in Table 1. Admission notes of hospitalized children were recorded in pre-designed forms in 274 (91.6%) of admittees, out of which 250 (83.6%) forms had a specific field for vaccination history. Admission notes contained DVH, "up-to-date vaccination" statement and no history of vaccination at all in 5 (1.7%), 175 (58.5%) and 119 (39.8%) charts, respectively. Bivariate analysis of the association of different variables with level of documentation is shown in Table 1.

*Fisher's Exact Test. † Out of 274 Clerking Forms. DVH= detailed vaccination history

Logistic regression showed no predictive effect of any of these variables for the level of documentation of vaccination history, the documentation of DVH and the absence of any documentation. Controlling for the name of the hospital did not affect the result of the analysis.
Only two admittees had a discharge diagnosis of AEFI. These were a case of BCGitis and post vaccination convulsion, respectively. Discharge diagnoses that the authors have labeled as "possibly-missed AEFI" are summarized in Table 2. None of their charts hold DVH .

This study demonstrated a very low documentation of DVH, which was less than 2% of admittees in the study period. Worth noting is that no specific factor could predict this behavior. This might be due to the low frequency of admittees with documented DVH, or to other factors that were not included in our study, such as practitioner's awareness.

However, DVH tends to be documented in clerking forms holding a vaccination field, which might work as a reminder. Earlier studies have shown that chart reminders and pre-printed forms have resulted in more complete and more succinct admission notes^9,10. DVH is found to be more in admittees less than one year of age, which clearly reflects the perception that most vaccination shots are accumulated in this age group.

No vaccination history was documented in high proportion (39.8%) of admittees. This is in comparison with 22% failure to document vaccination status reported by Philip Kum-Nji and colleagues in Tennessee. Low levels of documentation of vaccination history is likely to result in missed recognition of AEFI and missed opportunity to catch missed immunizations for possible catch-up correction.

Interesting to note is that all admittees with no clerking forms and all clerking forms with no pre-defined vaccination field have no documentation of vaccination history. This would emphasize the role of such forms and their design in directing the practitioners towards better documentation.

None of the likely AEFI had DVH. This might reflect the low recognition of practitioners towards AEFI.
Expected post-vaccination admissions in studied districts have been calculated using the local vaccination data and an attributable risk of admission in vaccinee less than 2 years of age of 22.5/100,000 vaccinee¹¹. The number of vaccinees in studied districts is almost 160,000¹². Thus the expected post-vaccination number of admissions in studied districts is 36 vaccinees. The expected number in our sample is 3 to 4 admittees. This study could elicit only one discharge diagnosis of an AEFI, i.e. less than one third of the expected number.

This low number, in addition to the very low documentation of DVH, suggests low recognition of this problem as a reason for hospital admission.

Documentation of vaccination history is poor and diagnosis of AEFI is low. It is possibly due to low recognition and inertia of practitioners. This poor practice might be a reason for the very poor reporting of AEFI.

We would like to thank all doctors and nurses working in the sampled hospitals for their contribution in collection of data. Particular thanks to Mr. Abdullah Abdulaal for his valuable secretarial work.